enlightenment/src/view.c

2871 lines
63 KiB
C
Raw Normal View History

#include <libefsd.h>
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
#include "debug.h"
#include "view.h"
#include "cursors.h"
#include "config.h"
#include "border.h"
#include "menu.h"
#include "menubuild.h"
#include "fs.h"
#include "file.h"
#include "util.h"
#include "globals.h"
static Ecore_Event *current_ev = NULL;
static char **dnd_files = NULL;
static int dnd_num_files = 0;
static E_dnd_enum dnd_pending_mode;
static void e_bg_down_cb(void *_data, Evas _e, Evas_Object _o, int _b, int _x, int _y);
static void e_bg_up_cb(void *_data, Evas _e, Evas_Object _o, int _b, int _x, int _y);
static void e_bg_move_cb(void *_data, Evas _e, Evas_Object _o, int _b, int _x, int _y);
static void e_idle(void *data);
static void e_wheel(Ecore_Event * ev);
static void e_key_down(Ecore_Event * ev);
static void e_key_up(Ecore_Event * ev);
static void e_mouse_down(Ecore_Event * ev);
static void e_mouse_up(Ecore_Event * ev);
static void e_mouse_move(Ecore_Event * ev);
static void e_mouse_in(Ecore_Event * ev);
static void e_mouse_out(Ecore_Event * ev);
static void e_window_expose(Ecore_Event * ev);
static void e_configure(Ecore_Event * ev);
static void e_property(Ecore_Event * ev);
static void e_unmap(Ecore_Event * ev);
static void e_visibility(Ecore_Event * ev);
static void e_focus_in(Ecore_Event * ev);
static void e_focus_out(Ecore_Event * ev);
static void e_delete(Ecore_Event * ev);
static void e_dnd_status(Ecore_Event * ev);
static void e_dnd_data_request(Ecore_Event * ev);
static void e_dnd_drop_end(Ecore_Event * ev);
static void e_dnd_drop_position(Ecore_Event * ev);
static void e_dnd_drop(Ecore_Event * ev);
static void e_dnd_drop_request(Ecore_Event * ev);
static void e_dnd_drop_request_free(void);
static void e_dnd_handle_drop( E_View *v, E_dnd_enum dnd_pending_mode );
static void e_view_handle_fs(EfsdEvent *ev);
static void e_view_handle_fs_restart(void *data);
static void e_view_resort_timeout(int val, void *data);
static int e_view_restart_alphabetical_qsort_cb(const void *data1, const void *data2);
static void e_view_geometry_record_timeout(int val, void *data);
static void e_view_scrollbar_v_change_cb(void *_data, E_Scrollbar *sb, double val);
static void e_view_scrollbar_h_change_cb(void *_data, E_Scrollbar *sb, double val);
static void e_view_write_icon_xy_timeout(int val, void *data);
static void e_view_bg_reload_timeout(int val, void *data);
static void
e_view_write_icon_xy_timeout(int val, void *data)
{
E_View *v;
Evas_List l;
E_Icon *ic;
double t_in;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
v = data;
/* FIXME: this is a problem if we have 1000's of icons */
t_in = ecore_get_time();
for (l = v->icons; l; l = l->next)
{
ic = l->data;
if (ic->q.write_xy)
{
char buf[PATH_MAX];
ic->q.write_xy = 0;
sprintf(buf, "%s/%s", ic->view->dir, ic->file);
D("write meta xy for icon for file %s\n", ic->file);
efsd_set_metadata_int(e_fs_get_connection(),
"/pos/x", buf,
ic->geom.x);
efsd_set_metadata_int(e_fs_get_connection(),
"/pos/y", buf,
ic->geom.y);
}
if ((ecore_get_time() - t_in) > 0.10)
{
char name[PATH_MAX];
sprintf(name, "icon_xy_record.%s", v->dir);
ecore_add_event_timer(name, 0.01, e_view_write_icon_xy_timeout, 0, v);
D_RETURN;
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
UN(val);
}
void
e_view_selection_update(E_View *v)
{
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
if ((v->select.on) && (!v->select.obj.middle))
{
Evas_Gradient grad;
/* create select objects */
v->select.obj.middle = evas_add_rectangle(v->evas);
evas_set_color(v->evas, v->select.obj.middle,
v->select.config.middle.r,
v->select.config.middle.g,
v->select.config.middle.b,
v->select.config.middle.a);
evas_set_layer(v->evas, v->select.obj.middle, 300);
v->select.obj.edge_l = evas_add_rectangle(v->evas);
evas_set_color(v->evas, v->select.obj.edge_l,
v->select.config.edge_l.r,
v->select.config.edge_l.g,
v->select.config.edge_l.b,
v->select.config.edge_l.a);
evas_set_layer(v->evas, v->select.obj.edge_l, 300);
v->select.obj.edge_r = evas_add_rectangle(v->evas);
evas_set_color(v->evas, v->select.obj.edge_r,
v->select.config.edge_r.r,
v->select.config.edge_r.g,
v->select.config.edge_r.b,
v->select.config.edge_r.a);
evas_set_layer(v->evas, v->select.obj.edge_r, 300);
v->select.obj.edge_t = evas_add_rectangle(v->evas);
evas_set_color(v->evas, v->select.obj.edge_t,
v->select.config.edge_t.r,
v->select.config.edge_t.g,
v->select.config.edge_t.b,
v->select.config.edge_t.a);
evas_set_layer(v->evas, v->select.obj.edge_t, 300);
v->select.obj.edge_b = evas_add_rectangle(v->evas);
evas_set_color(v->evas, v->select.obj.edge_b,
v->select.config.edge_b.r,
v->select.config.edge_b.g,
v->select.config.edge_b.b,
v->select.config.edge_b.a);
evas_set_layer(v->evas, v->select.obj.edge_b, 300);
v->select.obj.grad_l = evas_add_gradient_box(v->evas);
evas_set_angle(v->evas, v->select.obj.grad_l, 270);
grad = evas_gradient_new();
evas_gradient_add_color(grad,
v->select.config.grad_l.r,
v->select.config.grad_l.g,
v->select.config.grad_l.b,
v->select.config.grad_l.a, 8);
evas_gradient_add_color(grad,
v->select.config.grad_l.r,
v->select.config.grad_l.g,
v->select.config.grad_l.b,
0, 8);
evas_set_gradient(v->evas, v->select.obj.grad_l, grad);
evas_gradient_free(grad);
evas_set_layer(v->evas, v->select.obj.grad_l, 300);
v->select.obj.grad_r = evas_add_gradient_box(v->evas);
evas_set_angle(v->evas, v->select.obj.grad_r, 90);
grad = evas_gradient_new();
evas_gradient_add_color(grad,
v->select.config.grad_r.r,
v->select.config.grad_r.g,
v->select.config.grad_r.b,
v->select.config.grad_r.a, 8);
evas_gradient_add_color(grad,
v->select.config.grad_r.r,
v->select.config.grad_r.g,
v->select.config.grad_r.b,
0, 8);
evas_set_gradient(v->evas, v->select.obj.grad_r, grad);
evas_gradient_free(grad);
evas_set_layer(v->evas, v->select.obj.grad_r, 300);
v->select.obj.grad_t = evas_add_gradient_box(v->evas);
evas_set_angle(v->evas, v->select.obj.grad_t, 0);
grad = evas_gradient_new();
evas_gradient_add_color(grad,
v->select.config.grad_t.r,
v->select.config.grad_t.g,
v->select.config.grad_t.b,
v->select.config.grad_t.a, 8);
evas_gradient_add_color(grad,
v->select.config.grad_t.r,
v->select.config.grad_t.g,
v->select.config.grad_t.b,
0, 8);
evas_set_gradient(v->evas, v->select.obj.grad_t, grad);
evas_gradient_free(grad);
evas_set_layer(v->evas, v->select.obj.grad_t, 300);
v->select.obj.grad_b = evas_add_gradient_box(v->evas);
evas_set_angle(v->evas, v->select.obj.grad_b, 180);
grad = evas_gradient_new();
evas_gradient_add_color(grad,
v->select.config.grad_b.r,
v->select.config.grad_b.g,
v->select.config.grad_b.b,
v->select.config.grad_b.a, 8);
evas_gradient_add_color(grad,
v->select.config.grad_b.r,
v->select.config.grad_b.g,
v->select.config.grad_b.b,
0, 8);
evas_set_gradient(v->evas, v->select.obj.grad_b, grad);
evas_gradient_free(grad);
evas_set_layer(v->evas, v->select.obj.grad_b, 300);
v->select.obj.clip = evas_add_rectangle(v->evas);
evas_set_color(v->evas, v->select.obj.clip, 255, 255, 255, 255);
evas_set_clip(v->evas, v->select.obj.grad_l, v->select.obj.clip);
evas_set_clip(v->evas, v->select.obj.grad_r, v->select.obj.clip);
evas_set_clip(v->evas, v->select.obj.grad_t, v->select.obj.clip);
evas_set_clip(v->evas, v->select.obj.grad_b, v->select.obj.clip);
}
if ((!v->select.on) && (v->select.obj.middle))
{
/* destroy select objects */
evas_del_object(v->evas, v->select.obj.middle);
evas_del_object(v->evas, v->select.obj.edge_l);
evas_del_object(v->evas, v->select.obj.edge_r);
evas_del_object(v->evas, v->select.obj.edge_t);
evas_del_object(v->evas, v->select.obj.edge_b);
evas_del_object(v->evas, v->select.obj.grad_l);
evas_del_object(v->evas, v->select.obj.grad_r);
evas_del_object(v->evas, v->select.obj.grad_t);
evas_del_object(v->evas, v->select.obj.grad_b);
evas_del_object(v->evas, v->select.obj.clip);
v->select.obj.middle = NULL;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
if (!v->select.on) D_RETURN;
/* move & resize select objects */
{
evas_move(v->evas, v->select.obj.edge_l, v->select.x, v->select.y + 1);
evas_resize(v->evas, v->select.obj.edge_l, 1, v->select.h - 1);
evas_move(v->evas, v->select.obj.edge_r, v->select.x + v->select.w - 1, v->select.y);
evas_resize(v->evas, v->select.obj.edge_r, 1, v->select.h - 1);
evas_move(v->evas, v->select.obj.edge_t, v->select.x, v->select.y);
evas_resize(v->evas, v->select.obj.edge_t, v->select.w - 1, 1);
evas_move(v->evas, v->select.obj.edge_b, v->select.x + 1, v->select.y + v->select.h - 1);
evas_resize(v->evas, v->select.obj.edge_b, v->select.w - 1, 1);
evas_move(v->evas, v->select.obj.middle, v->select.x + 1, v->select.y + 1);
evas_resize(v->evas, v->select.obj.middle, v->select.w - 1 - 1, v->select.h - 1 - 1);
evas_move(v->evas, v->select.obj.grad_l, v->select.x + 1, v->select.y + 1);
evas_resize(v->evas, v->select.obj.grad_l, v->select.config.grad_size.l, v->select.h - 1 - 1);
evas_move(v->evas, v->select.obj.grad_r, v->select.x + v->select.w - 1 - v->select.config.grad_size.r, v->select.y + 1);
evas_resize(v->evas, v->select.obj.grad_r, v->select.config.grad_size.r, v->select.h - 1 - 1);
evas_move(v->evas, v->select.obj.grad_t, v->select.x + 1, v->select.y + 1);
evas_resize(v->evas, v->select.obj.grad_t, v->select.w - 1 - 1, v->select.config.grad_size.t);
evas_move(v->evas, v->select.obj.grad_b, v->select.x + 1, v->select.y + v->select.h - 1 - v->select.config.grad_size.b);
evas_resize(v->evas, v->select.obj.grad_b, v->select.w - 1 - 1, v->select.config.grad_size.b);
evas_move(v->evas, v->select.obj.clip, v->select.x + 1, v->select.y + 1);
evas_resize(v->evas, v->select.obj.clip, v->select.w - 1 - 1, v->select.h - 1 - 1);
}
evas_show(v->evas, v->select.obj.middle);
evas_show(v->evas, v->select.obj.edge_l);
evas_show(v->evas, v->select.obj.edge_r);
evas_show(v->evas, v->select.obj.edge_t);
evas_show(v->evas, v->select.obj.edge_b);
evas_show(v->evas, v->select.obj.grad_l);
evas_show(v->evas, v->select.obj.grad_r);
evas_show(v->evas, v->select.obj.grad_t);
evas_show(v->evas, v->select.obj.grad_b);
evas_show(v->evas, v->select.obj.clip);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_view_scrollbar_v_change_cb(void *_data, E_Scrollbar *sb, double val)
{
E_View *v;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
v = _data;
e_view_scroll_to(v, v->scroll.x, v->spacing.window.t - sb->val);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
UN(val);
}
static void
e_view_scrollbar_h_change_cb(void *_data, E_Scrollbar *sb, double val)
{
E_View *v;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
v = _data;
e_view_scroll_to(v, v->spacing.window.l - sb->val, v->scroll.y);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
UN(val);
}
static void
e_bg_down_cb(void *_data, Evas _e, Evas_Object _o, int _b, int _x, int _y)
{
Ecore_Event_Mouse_Down *ev;
E_View *v;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
if (!current_ev) D_RETURN;
ev = current_ev->event;
v = _data;
if (!(ev->mods & (multi_select_mod | range_select_mod)))
{
v->select.last_count = v->select.count;
e_view_deselect_all();
}
if (_b == 1)
{
v->select.down.x = _x;
v->select.down.y = _y;
v->select.on = 1;
if (_x < v->select.down.x)
{
v->select.x = _x;
v->select.w = v->select.down.x - v->select.x + 1;
}
else
{
v->select.x = v->select.down.x;
v->select.w = _x - v->select.down.x + 1;
}
if (_y < v->select.down.y)
{
v->select.y = _y;
v->select.h = v->select.down.y - v->select.y + 1;
}
else
{
v->select.y = v->select.down.y;
v->select.h = _y - v->select.down.y + 1;
}
e_view_selection_update(v);
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
if( _b == 2 && ev->double_click )
ecore_event_loop_quit();
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
UN(_e);
UN(_o);
}
static void
e_bg_up_cb(void *_data, Evas _e, Evas_Object _o, int _b, int _x, int _y)
{
Ecore_Event_Mouse_Up *ev;
E_View *v;
int dx, dy;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
if (!current_ev) D_RETURN;
ev = current_ev->event;
v = _data;
dx = 0;
dy = 0;
if (v->select.on)
{
dx = v->select.down.x - _x;
dy = v->select.down.y - _y;
if (dx < 0) dx = -dx;
if (dy < 0) dy = -dy;
if (_b == 1)
v->select.on = 0;
e_view_selection_update(v);
}
if ((_b == 1) && ((dx > 3) || (dy > 3)))
{
Evas_List l;
for (l = v->icons; l; l = l->next)
{
E_Icon *ic;
ic = l->data;
if (INTERSECTS(v->select.x, v->select.y, v->select.w, v->select.h,
v->scroll.x + ic->geom.x,
v->scroll.y + ic->geom.y, ic->geom.w, ic->geom.h))
{
if (ic->state.visible)
{
e_icon_invert_selection(ic);
}
}
}
}
else if (v->select.last_count == 0)
{
if (_b == 1)
{
if (!(ev->mods & (multi_select_mod | range_select_mod)))
{
static E_Build_Menu *buildmenu = NULL;
if (!buildmenu)
{
char *apps_menu_db;
apps_menu_db = e_config_get("apps_menu");
if (apps_menu_db) buildmenu = e_build_menu_new_from_db(apps_menu_db);
}
if (buildmenu)
{
static E_Menu *menu = NULL;
menu = buildmenu->menu;
if (menu)
2001-09-25 09:37:00 -07:00
{
e_menu_show_at_mouse(menu, ev->rx, ev->ry, ev->time);
}
}
}
}
/* else if (_b == 2)
{
exit(0);
}*/
else if (_b == 3)
{
static E_Build_Menu *buildmenu = NULL;
if (!buildmenu)
{
buildmenu = e_build_menu_new_from_gnome_apps("/usr/share/gnome/apps");
}
if (buildmenu)
{
static E_Menu *menu = NULL;
menu = buildmenu->menu;
if (menu)
e_menu_show_at_mouse(menu, ev->rx, ev->ry, ev->time);
}
}
}
if (_b == 1)
{
v->select.x = _x;
v->select.y = _y;
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
UN(_e);
UN(_o);
}
static void
e_bg_move_cb(void *_data, Evas _e, Evas_Object _o, int _b, int _x, int _y)
{
Ecore_Event_Mouse_Down *ev;
E_View *v;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
if (!current_ev) D_RETURN;
ev = current_ev->event;
v = _data;
if (v->select.on)
{
if (_x < v->select.down.x)
{
v->select.x = _x;
v->select.w = v->select.down.x - v->select.x + 1;
}
else
{
v->select.x = v->select.down.x;
v->select.w = _x - v->select.down.x + 1;
}
if (_y < v->select.down.y)
{
v->select.y = _y;
v->select.h = v->select.down.y - v->select.y + 1;
}
else
{
v->select.y = v->select.down.y;
v->select.h = _y - v->select.down.y + 1;
}
e_view_selection_update(v);
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
UN(_e);
UN(_o);
UN(_b);
}
void
e_view_deselect_all(void)
{
Evas_List ll;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
for (ll = views; ll; ll = ll->next)
{
Evas_List l;
E_View *v;
v = ll->data;
for (l = v->icons; l; l = l->next)
{
E_Icon *ic;
ic = l->data;
e_icon_deselect(ic);
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_deselect_all_except(E_Icon *not_ic)
{
Evas_List ll;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
for (ll = views; ll; ll = ll->next)
{
Evas_List l;
E_View *v;
v = ll->data;
for (l = v->icons; l; l = l->next)
{
E_Icon *ic;
ic = l->data;
if (ic != not_ic)
e_icon_deselect(ic);
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_icons_get_extents(E_View *v, int *min_x, int *min_y, int *max_x, int *max_y)
{
Evas_List l;
int x1, x2, y1, y2;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
x1 = v->extents.x1;
x2 = v->extents.x2;
y1 = v->extents.y1;
y2 = v->extents.y2;
if (!v->extents.valid)
{
x1 = 999999999;
x2 = -999999999;
y1 = 999999999;
y2 = -999999999;
if (!v->icons)
{
if (min_x) *min_x = 0;
if (min_y) *min_y = 0;
if (max_x) *max_x = 1;
if (max_y) *max_y = 1;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
for (l = v->icons; l; l = l->next)
{
E_Icon *ic;
ic = l->data;
if (ic->geom.x < x1) x1 = ic->geom.x;
if (ic->geom.y < y1) y1 = ic->geom.y;
if (ic->geom.x + ic->geom.w > x2) x2 = ic->geom.x + ic->geom.w;
if (ic->geom.y + ic->geom.h > y2) y2 = ic->geom.y + ic->geom.h;
}
v->extents.x1 = x1;
v->extents.y1 = y1;
v->extents.x2 = x2;
v->extents.y2 = y2;
}
v->extents.valid = 1;
if (x1 > 0) x1 = 0;
if (y1 > 0) y1 = 0;
if (min_x) *min_x = x1;
if (min_y) *min_y = y1;
if (max_x) *max_x = x2 - 1;
if (max_y) *max_y = y2 - 1;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_icons_apply_xy(E_View *v)
{
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
for (l = v->icons; l; l = l->next)
{
E_Icon *ic;
ic = l->data;
e_icon_apply_xy(ic);
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_scroll_to(E_View *v, int sx, int sy)
{
int min_x, min_y, max_x, max_y;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e_view_icons_get_extents(v, &min_x, &min_y, &max_x, &max_y);
if (sx < v->size.w - v->spacing.window.r - max_x)
sx = v->size.w - v->spacing.window.r - max_x;
if (sx > v->spacing.window.l - min_x)
sx = v->spacing.window.l - min_x;
if (sy < v->size.h - v->spacing.window.b - max_y)
sy = v->size.h - v->spacing.window.b - max_y;
if (sy > v->spacing.window.t - min_y)
sy = v->spacing.window.t - min_y;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
if ((sx == v->scroll.x) && (v->scroll.y == sy)) D_RETURN;
v->scroll.x = sx;
v->scroll.y = sy;
e_view_icons_apply_xy(v);
if (v->bg) e_background_set_scroll(v->bg, v->scroll.x, v->scroll.y);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_scroll_by(E_View *v, int sx, int sy)
{
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e_view_scroll_to(v, v->scroll.x + sx, v->scroll.y + sy);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_scroll_to_percent(E_View *v, double psx, double psy)
{
int min_x, min_y, max_x, max_y;
int sx, sy;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e_view_icons_get_extents(v, &min_x, &min_y, &max_x, &max_y);
sx = (psx * ((double)max_x - (double)min_x)) - min_x;
sy = (psy * ((double)max_y - (double)min_y)) - min_y;
if (sx < v->size.w - v->spacing.window.r - max_x)
sx = v->size.w - v->spacing.window.r - max_x;
if (sx > v->spacing.window.l - min_x)
sx = v->spacing.window.l - min_x;
if (sy < v->size.h - v->spacing.window.b - max_y)
sy = v->size.h - v->spacing.window.b - max_y;
if (sy > v->spacing.window.t - min_y)
sy = v->spacing.window.t - min_y;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
if ((sx == v->scroll.x) && (v->scroll.y == sy)) D_RETURN;
v->scroll.x = sx;
v->scroll.y = sy;
e_view_icons_apply_xy(v);
if (v->bg) e_background_set_scroll(v->bg, v->scroll.x, v->scroll.y);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_get_viewable_percentage(E_View *v, double *vw, double *vh)
{
int min_x, min_y, max_x, max_y;
double p;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e_view_icons_get_extents(v, &min_x, &min_y, &max_x, &max_y);
if (min_x == max_x)
{
if (vw) *vw = 0;
}
else
{
p = ((double)(v->size.w - v->spacing.window.l - v->spacing.window.r)) /
((double)(max_x - min_x));
if (vw) *vw = p;
}
if (min_y == max_y)
{
if (vh) *vh = 0;
}
else
{
p = ((double)(v->size.h - v->spacing.window.t - v->spacing.window.b)) /
((double)(max_y - min_y));
if (vh) *vh = p;
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_get_position_percentage(E_View *v, double *vx, double *vy)
{
int min_x, min_y, max_x, max_y;
double p;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e_view_icons_get_extents(v, &min_x, &min_y, &max_x, &max_y);
if (min_x == max_x)
{
if (vx) *vx = 0;
}
else
{
p = ((double)(v->scroll.x - min_x)) / ((double)(max_x - min_x));
if (vx) *vx = p;
}
if (min_y == max_y)
{
if (vy) *vy = 0;
}
else
{
p = ((double)(v->scroll.y - min_y)) / ((double)(max_y - min_y));
if (vy) *vy = p;
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_idle(void *data)
{
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
e_view_update(v);
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
UN(data);
}
void
e_view_geometry_record(E_View *v)
{
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
if (e_fs_get_connection())
{
int left, top;
D("Record geom for view\n");
ecore_window_get_frame_size(v->win.base, &left, NULL,
&top, NULL);
efsd_set_metadata_int(e_fs_get_connection(),
"/view/x", v->dir,
v->location.x - left);
efsd_set_metadata_int(e_fs_get_connection(),
"/view/y", v->dir,
v->location.y - top);
efsd_set_metadata_int(e_fs_get_connection(),
"/view/w", v->dir,
v->size.w);
efsd_set_metadata_int(e_fs_get_connection(),
"/view/h", v->dir,
v->size.h);
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_view_geometry_record_timeout(int val, void *data)
{
E_View *v;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
v = data;
e_view_geometry_record(v);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
UN(val);
}
void
e_view_queue_geometry_record(E_View *v)
{
char name[PATH_MAX];
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
sprintf(name, "geometry_record.%s", v->dir);
ecore_add_event_timer(name, 0.10, e_view_geometry_record_timeout, 0, v);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_queue_icon_xy_record(E_View *v)
{
char name[PATH_MAX];
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
sprintf(name, "icon_xy_record.%s", v->dir);
ecore_add_event_timer(name, 0.10, e_view_write_icon_xy_timeout, 0, v);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_configure(Ecore_Event * ev)
{
Ecore_Event_Window_Configure *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.base)
{
/* win, root, x, y, w, h, wm_generated */
D("configure for view %s\n", v->dir);
if (e->wm_generated)
{
D("wm generated %i %i, %ix%i\n", e->x, e->y, e->w, e->h);
if ((e->x != v->location.x) || (e->y != v->location.y))
{
D("new spot!\n");
v->location.x = e->x;
v->location.y = e->y;
e_view_queue_geometry_record(v);
}
}
D("size %ix%i\n", e->w, e->h);
if ((e->w != v->size.w) || (e->h != v->size.h) || (v->size.force))
{
v->size.force = 0;
D("... a new size!\n");
v->size.w = e->w;
v->size.h = e->h;
if (v->pmap) ecore_pixmap_free(v->pmap);
v->pmap = 0;
ecore_window_resize(v->win.main, v->size.w, v->size.h);
if (v->options.back_pixmap)
{
v->pmap = ecore_pixmap_new(v->win.main, v->size.w, v->size.h, 0);
evas_set_output(v->evas, ecore_display_get(), v->pmap,
evas_get_visual(v->evas),
evas_get_colormap(v->evas));
ecore_window_set_background_pixmap(v->win.main, v->pmap);
ecore_window_clear(v->win.main);
}
if (v->bg) e_background_set_size(v->bg, v->size.w, v->size.h);
D("evas_set_output_viewpor(%p)\n", v->evas);
evas_set_output_viewport(v->evas, 0, 0, v->size.w, v->size.h);
evas_set_output_size(v->evas, v->size.w, v->size.h);
e_view_scroll_to(v, v->scroll.x, v->scroll.y);
e_view_arrange(v);
e_view_queue_geometry_record(v);
e_scrollbar_move(v->scrollbar.v, v->size.w - 12, 0);
e_scrollbar_resize(v->scrollbar.v, 12, v->size.h - 12);
e_scrollbar_move(v->scrollbar.h, 0, v->size.h - 12);
e_scrollbar_resize(v->scrollbar.h, v->size.w - 12, 12);
if (v->iconbar) e_iconbar_fix(v->iconbar);
}
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_property(Ecore_Event * ev)
{
Ecore_Event_Window_Configure *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.base)
{
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_unmap(Ecore_Event * ev)
{
Ecore_Event_Window_Unmap *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.base)
{
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_visibility(Ecore_Event * ev)
{
Ecore_Event_Window_Unmap *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.base)
{
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_focus_in(Ecore_Event * ev)
{
Ecore_Event_Window_Focus_In *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.base)
{
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_focus_out(Ecore_Event * ev)
{
Ecore_Event_Window_Focus_Out *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.base)
{
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_delete(Ecore_Event * ev)
{
Ecore_Event_Window_Delete *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.base)
{
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
e_object_unref(E_OBJECT(v));
D_RETURN;
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
/*
* dnd status handler
*
*/
static void
e_dnd_status(Ecore_Event * ev)
{
Ecore_Event_Dnd_Drop_Status *e;
/*
typedef struct _ecore_event_dnd_drop_status
{
Window win, root, source_win;
int x, y, w, h;
int ok;
} Ecore_Event_Dnd_Drop_Status;
*/
Evas_List l;
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.base)
{
ecore_window_dnd_ok(e->ok);
}
}
D_RETURN;
}
static void
e_wheel(Ecore_Event * ev)
{
Ecore_Event_Wheel *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.main)
{
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_key_down(Ecore_Event * ev)
{
Ecore_Event_Key_Down *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if ((e->win == v->win.base) ||
(e->win == v->win.main))
{
if (!strcmp(e->key, "Up"))
{
e_scrollbar_set_value(v->scrollbar.v, v->scrollbar.v->val - 8);
}
else if (!strcmp(e->key, "Down"))
{
e_scrollbar_set_value(v->scrollbar.v, v->scrollbar.v->val + 8);
}
else if (!strcmp(e->key, "Left"))
{
e_scrollbar_set_value(v->scrollbar.h, v->scrollbar.h->val - 8);
}
else if (!strcmp(e->key, "Right"))
{
e_scrollbar_set_value(v->scrollbar.h, v->scrollbar.h->val + 8);
}
else if (!strcmp(e->key, "Escape"))
{
}
else
{
char *type;
type = ecore_keypress_translate_into_typeable(e);
if (type)
{
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_key_up(Ecore_Event * ev)
{
Ecore_Event_Key_Up *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_mouse_down(Ecore_Event * ev)
{
Ecore_Event_Mouse_Down *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
current_ev = ev;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.main)
{
int focus_mode;
E_CFG_INT(cfg_focus_mode, "settings", "/focus/mode", 0);
E_CONFIG_INT_GET(cfg_focus_mode, focus_mode);
if (focus_mode == 2)
ecore_focus_to_window(v->win.base);
evas_event_button_down(v->evas, e->x, e->y, e->button);
current_ev = NULL;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
}
current_ev = NULL;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_mouse_up(Ecore_Event * ev)
{
Ecore_Event_Mouse_Up *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
current_ev = ev;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.main)
{
evas_event_button_up(v->evas, e->x, e->y, e->button);
current_ev = NULL;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
}
current_ev = NULL;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_mouse_move(Ecore_Event * ev)
{
Ecore_Event_Mouse_Move *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
current_ev = ev;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.main)
{
evas_event_move(v->evas, e->x, e->y);
current_ev = NULL;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
}
current_ev = NULL;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_mouse_in(Ecore_Event * ev)
{
Ecore_Event_Window_Enter *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
current_ev = ev;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.main)
{
if (v->is_desktop)
{
evas_event_enter(v->evas);
}
current_ev = NULL;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
}
current_ev = NULL;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_mouse_out(Ecore_Event * ev)
{
Ecore_Event_Window_Leave *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
current_ev = ev;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.main)
{
evas_event_leave(v->evas);
current_ev = NULL;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
}
current_ev = NULL;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_window_expose(Ecore_Event * ev)
{
Ecore_Event_Window_Expose *e;
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.main)
{
if (!(v->pmap))
evas_update_rect(v->evas, e->x, e->y, e->w, e->h);
v->changed = 1;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_view_handle_fs_restart(void *data)
{
E_View *v;
Evas_List icons = NULL, l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
v = data;
D("e_view_handle_fs_restart\n");
for (l = v->icons; l; l = l->next)
{
icons = evas_list_prepend(icons, l->data);
}
if (icons)
{
for (l = icons; l; l = l->next)
{
E_Icon *i;
i = l->data;
e_view_file_deleted(v->monitor_id, i->file);
}
evas_list_free(icons);
}
if (e_fs_get_connection())
{
if (v->geom_get.busy)
{
v->geom_get.x = efsd_get_metadata(e_fs_get_connection(),
"/view/x", v->dir, EFSD_INT);
v->geom_get.y = efsd_get_metadata(e_fs_get_connection(),
"/view/y", v->dir, EFSD_INT);
v->geom_get.w = efsd_get_metadata(e_fs_get_connection(),
"/view/w", v->dir, EFSD_INT);
v->geom_get.h = efsd_get_metadata(e_fs_get_connection(),
"/view/h", v->dir, EFSD_INT);
v->getbg = efsd_get_metadata(e_fs_get_connection(),
"/view/background", v->dir, EFSD_STRING);
}
{
EfsdOptions *ops;
ops = efsd_ops(3,
efsd_op_get_stat(),
efsd_op_get_filetype(),
efsd_op_list_all());
v->monitor_id = efsd_start_monitor(e_fs_get_connection(), v->dir,
ops, TRUE);
2001-08-16 17:57:41 -07:00
}
v->is_listing = 1;
}
D("restarted monitor id (connection = %p), %i for %s\n", e_fs_get_connection(), v->monitor_id, v->dir);
v->is_listing = 1;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
Ecore_Event *
e_view_get_current_event(void)
{
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
D_RETURN_(current_ev);
}
int
e_view_filter_file(E_View *v, char *file)
{
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
if (file[0] == '.')
D_RETURN_(0);
D_RETURN_(1);
UN(v);
}
static int
e_view_restart_alphabetical_qsort_cb(const void *data1, const void *data2)
{
E_Icon *ic, *ic2;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
ic = *((E_Icon **)data1);
ic2 = *((E_Icon **)data2);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN_(strcmp(ic->file, ic2->file));
}
void
e_view_resort_alphabetical(E_View *v)
{
Evas_List icons = NULL, l;
E_Icon **array;
int i, count;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
if (!v->icons) D_RETURN;
for (count = 0, l = v->icons; l; l = l->next) count++;
array = malloc(sizeof(E_Icon *) * count);
for (i = 0, l = v->icons; l; l = l->next) array[i++] = l->data;
D("qsort %i elements...\n", count);
qsort(array, count, sizeof(E_Icon *),
e_view_restart_alphabetical_qsort_cb);
for (i = 0; i < count; i++) icons = evas_list_append(icons, array[i]);
free(array);
evas_list_free(v->icons);
v->icons = icons;
D("done...\n");
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_arrange(E_View *v)
{
Evas_List l;
int x, y;
int x1, x2, y1, y2;
double sv, sr, sm;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
x = v->spacing.window.l;
y = v->spacing.window.t;
for (l = v->icons; l; l = l->next)
{
E_Icon *ic;
ic = l->data;
if ((x != v->spacing.window.l) &&
((x + ic->geom.w) > v->size.w - v->spacing.window.r))
{
x = v->spacing.window.l;
y += ic->geom.h + v->spacing.icon.b;
}
ic->geom.x = x;
ic->geom.y = y;
e_icon_apply_xy(ic);
x += ic->geom.w + v->spacing.icon.s;
}
e_view_icons_get_extents(v, &x1, &y1, &x2, &y2);
sv = - (v->scroll.y - v->spacing.window.t);
sr = v->size.h - v->spacing.window.t - v->spacing.window.b;
sm = y2 - y1;
if (sr > sm) sr = sm;
e_scrollbar_set_value(v->scrollbar.v, sv);
e_scrollbar_set_range(v->scrollbar.v, sr);
e_scrollbar_set_max(v->scrollbar.v, sm);
if (sr < sm) e_scrollbar_show(v->scrollbar.v);
else e_scrollbar_hide(v->scrollbar.v);
sv = - (v->scroll.x - v->spacing.window.l);
sr = v->size.w - v->spacing.window.l - v->spacing.window.r;
sm = x2 - x1;
if (sr > sm) sr = sm;
e_scrollbar_set_value(v->scrollbar.h, sv);
e_scrollbar_set_range(v->scrollbar.h, sr);
e_scrollbar_set_max(v->scrollbar.h, sm);
if (sr < sm) e_scrollbar_show(v->scrollbar.h);
else e_scrollbar_hide(v->scrollbar.h);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_resort(E_View *v)
{
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
e_view_resort_alphabetical(v);
e_view_arrange(v);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_view_resort_timeout(int val, void *data)
{
E_View *v;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
v = data;
e_view_resort(v);
v->have_resort_queued = 0;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
UN(val);
}
void
e_view_queue_resort(E_View *v)
{
char name[PATH_MAX];
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
if (v->have_resort_queued) D_RETURN;
v->have_resort_queued = 1;
sprintf(name, "resort_timer.%s", v->dir);
ecore_add_event_timer(name, 1.0, e_view_resort_timeout, 0, v);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_file_added(int id, char *file)
{
E_View *v;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
/* if we get a path - ignore it - its not a file in the dir */
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
if (!file) D_RETURN;
/* D("FILE ADD: %s\n", file);*/
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
if (file[0] == '/') D_RETURN;
v = e_view_find_by_monitor_id(id);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
if (!v) D_RETURN;
e_iconbar_file_add(v, file);
e_view_bg_add(v, file);
/* filter files here */
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
if (!e_view_filter_file(v, file)) D_RETURN;
if (!e_icon_find_by_file(v, file))
{
E_Icon *ic;
ic = e_icon_new();
ic->view = v;
ic->file = strdup(file);
ic->changed = 1;
ic->obj.icon = evas_add_image_from_file(ic->view->evas, NULL);
ic->obj.text = e_text_new(ic->view->evas, ic->file, "filename");
v->icons = evas_list_append(v->icons, ic);
v->extents.valid = 0;
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_file_deleted(int id, char *file)
{
E_View *v;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
if (!file) D_RETURN;
if (file[0] == '/') D_RETURN;
v = e_view_find_by_monitor_id(id);
2001-11-16 20:21:48 -08:00
if (!v) D_RETURN;
e_iconbar_file_delete(v, file);
e_view_bg_del(v, file);
{
E_Icon *ic;
ic = e_icon_find_by_file(v, file);
if (ic)
{
e_icon_hide(ic);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
e_object_unref(E_OBJECT(ic));
v->icons = evas_list_remove(v->icons, ic);
v->changed = 1;
v->extents.valid = 0;
e_view_queue_resort(v);
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_file_changed(int id, char *file)
{
E_View *v;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
if (!file) D_RETURN;
if (file[0] == '/') D_RETURN;
v = e_view_find_by_monitor_id(id);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
if (!v) D_RETURN;
e_iconbar_file_change(v, file);
e_view_bg_change(v, file);
{
E_Icon *ic;
ic = e_icon_find_by_file(v, file);
if (ic)
{
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_file_moved(int id, char *file)
{
E_View *v;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
/* never gets called ? */
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
if (!file) D_RETURN;
D(".!WOW!. e_view_file_moved(%i, %s);\n", id, file);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
if (file[0] == '/') D_RETURN;
v = e_view_find_by_monitor_id(id);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
if (!v) D_RETURN;
{
E_Icon *ic;
ic = e_icon_find_by_file(v, file);
if (ic)
{
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
E_View *
e_view_find_by_monitor_id(int id)
{
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
if (v->monitor_id == id)
D_RETURN_(v);
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN_(NULL);
}
void
e_view_close_all(void)
{
while (views)
{
E_View *v;
v = views->data;
e_object_unref(E_OBJECT(v));
}
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
static void
e_view_cleanup(E_View *v)
{
char name[PATH_MAX];
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
if (v->iconbar)
{
e_iconbar_save_out_final(v->iconbar);
e_object_unref(E_OBJECT(v->iconbar));
}
sprintf(name, "resort_timer.%s", v->dir);
ecore_del_event_timer(name);
sprintf(name, "geometry_record.%s", v->dir);
ecore_del_event_timer(name);
sprintf(name, "icon_xy_record.%s", v->dir);
ecore_del_event_timer(name);
views = evas_list_remove(views, v);
2001-07-29 23:21:28 -07:00
efsd_stop_monitor(e_fs_get_connection(), v->dir, TRUE);
if (v->restarter)
e_fs_del_restart_handler(v->restarter);
v->restarter = NULL;
ecore_window_destroy(v->win.base);
/* FIXME: clean up the rest!!! this leaks ... */
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
/* Call the destructor of the base class */
e_object_cleanup(E_OBJECT(v));
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
E_View *
e_view_new(void)
{
E_View *v;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
v = NEW(E_View, 1);
ZERO(v, E_View, 1);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
e_object_init(E_OBJECT(v), (E_Cleanup_Func) e_view_cleanup);
#define SOFT_DESK
/* #define X_DESK */
/* #define GL_DESK */
#ifdef SOFT_DESK
/* ONLY alpha software can be "backing stored" */
v->options.render_method = RENDER_METHOD_ALPHA_SOFTWARE;
v->options.back_pixmap = 1;
#else
#ifdef X_DESK
v->options.render_method = RENDER_METHOD_BASIC_HARDWARE;
v->options.back_pixmap = 0;
#else
v->options.render_method = RENDER_METHOD_3D_HARDWARE;
v->options.back_pixmap = 0;
#endif
#endif
v->select.config.grad_size.l = 8;
v->select.config.grad_size.r = 8;
v->select.config.grad_size.t = 8;
v->select.config.grad_size.b = 8;
#define SETCOL(_member, _r, _g, _b, _a) \
_member.r = _r; _member.g = _g; _member.b = _b; _member.a = _a;
SETCOL(v->select.config.edge_l, 0, 0, 0, 255);
SETCOL(v->select.config.edge_r, 0, 0, 0, 255);
SETCOL(v->select.config.edge_t, 0, 0, 0, 255);
SETCOL(v->select.config.edge_b, 0, 0, 0, 255);
SETCOL(v->select.config.middle, 255, 255, 255, 100);
SETCOL(v->select.config.grad_l, 255, 255, 255, 100);
SETCOL(v->select.config.grad_r, 255, 255, 255, 100);
SETCOL(v->select.config.grad_t, 255, 255, 255, 100);
SETCOL(v->select.config.grad_b, 255, 255, 255, 100);
2001-03-15 17:14:42 -08:00
v->spacing.window.l = 3;
v->spacing.window.r = 15;
v->spacing.window.t = 3;
v->spacing.window.b = 15;
2001-08-01 22:08:49 -07:00
v->spacing.icon.s = 7;
v->spacing.icon.g = 7;
v->spacing.icon.b = 7;
views = evas_list_append(views, v);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN_(v);
}
void
e_view_set_background(E_View *v)
{
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
v->changed = 1;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_set_dir(E_View *v, char *dir)
{
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
/* stop monitoring old dir */
if ((v->dir) && (v->monitor_id))
{
efsd_stop_monitor(e_fs_get_connection(), v->dir, TRUE);
v->monitor_id = 0;
}
IF_FREE(v->dir);
v->dir = e_file_realpath(dir);
/* start monitoring new dir */
v->restarter = e_fs_add_restart_handler(e_view_handle_fs_restart, v);
if (e_fs_get_connection())
{
v->geom_get.x = efsd_get_metadata(e_fs_get_connection(),
"/view/x", v->dir, EFSD_INT);
v->geom_get.y = efsd_get_metadata(e_fs_get_connection(),
"/view/y", v->dir, EFSD_INT);
v->geom_get.w = efsd_get_metadata(e_fs_get_connection(),
"/view/w", v->dir, EFSD_INT);
v->geom_get.h = efsd_get_metadata(e_fs_get_connection(),
"/view/h", v->dir, EFSD_INT);
v->getbg = efsd_get_metadata(e_fs_get_connection(),
"/view/background", v->dir, EFSD_STRING);
v->geom_get.busy = 1;
{
EfsdOptions *ops;
ops = efsd_ops(3,
efsd_op_get_stat(),
efsd_op_get_filetype(),
efsd_op_list_all());
v->monitor_id = efsd_start_monitor(e_fs_get_connection(), v->dir,
ops, TRUE);
}
D("monitor id for %s = %i\n", v->dir, v->monitor_id);
v->is_listing = 1;
v->changed = 1;
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_realize(E_View *v)
{
int max_colors = 216;
int font_cache = 1024 * 1024;
int image_cache = 8192 * 1024;
char *font_dir;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
if (v->evas) D_RETURN;
v->win.base = ecore_window_new(0,
v->location.x, v->location.y,
v->size.w, v->size.h);
ecore_window_set_delete_inform(v->win.base);
font_dir = e_config_get("fonts");
v->evas = evas_new_all(ecore_display_get(),
v->win.base,
0, 0, v->size.w, v->size.h,
v->options.render_method,
max_colors,
font_cache,
image_cache,
font_dir);
v->win.main = evas_get_window(v->evas);
e_cursors_display_in_window(v->win.main, "View");
2001-03-15 17:14:42 -08:00
evas_event_move(v->evas, -999999, -999999);
ecore_window_set_events(v->win.base,
XEV_VISIBILITY | XEV_CONFIGURE |
XEV_PROPERTY | XEV_FOCUS);
ecore_window_set_events(v->win.main,
XEV_EXPOSE | XEV_MOUSE_MOVE |
XEV_BUTTON | XEV_IN_OUT | XEV_KEY);
if (v->options.back_pixmap)
{
v->pmap = ecore_pixmap_new(v->win.main, v->size.w, v->size.h, 0);
evas_set_output(v->evas, ecore_display_get(), v->pmap,
evas_get_visual(v->evas),
evas_get_colormap(v->evas));
ecore_window_set_background_pixmap(v->win.main, v->pmap);
}
if (v->bg)
{
e_background_realize(v->bg, v->evas);
e_background_set_size(v->bg, v->size.w, v->size.h);
}
v->obj_bg = evas_add_rectangle(v->evas);
evas_callback_add(v->evas, v->obj_bg, CALLBACK_MOUSE_DOWN, e_bg_down_cb, v);
evas_callback_add(v->evas, v->obj_bg, CALLBACK_MOUSE_UP, e_bg_up_cb, v);
evas_callback_add(v->evas, v->obj_bg, CALLBACK_MOUSE_MOVE, e_bg_move_cb, v);
evas_set_layer(v->evas, v->obj_bg, 100);
evas_move(v->evas, v->obj_bg, 0, 0);
evas_resize(v->evas, v->obj_bg, 999999, 999999);
evas_set_color(v->evas, v->obj_bg, 0, 0, 0, 0);
evas_show(v->evas, v->obj_bg);
v->scrollbar.v = e_scrollbar_new();
e_scrollbar_set_change_func(v->scrollbar.v, e_view_scrollbar_v_change_cb, v);
e_scrollbar_set_direction(v->scrollbar.v, 1);
e_scrollbar_add_to_evas(v->scrollbar.v, v->evas);
e_scrollbar_set_layer(v->scrollbar.v, 2000);
e_scrollbar_set_value(v->scrollbar.v, 0.0);
e_scrollbar_set_range(v->scrollbar.v, 1.0);
e_scrollbar_set_max(v->scrollbar.v, 1.0);
v->scrollbar.h = e_scrollbar_new();
e_scrollbar_set_change_func(v->scrollbar.h, e_view_scrollbar_h_change_cb, v);
e_scrollbar_set_direction(v->scrollbar.h, 0);
e_scrollbar_add_to_evas(v->scrollbar.h, v->evas);
e_scrollbar_set_layer(v->scrollbar.h, 2000);
e_scrollbar_set_value(v->scrollbar.h, 0.0);
e_scrollbar_set_range(v->scrollbar.h, 1.0);
e_scrollbar_set_max(v->scrollbar.h, 1.0);
e_scrollbar_move(v->scrollbar.v, v->size.w - 12, 0);
e_scrollbar_resize(v->scrollbar.v, 12, v->size.h - 12);
e_scrollbar_move(v->scrollbar.h, 0, v->size.h - 12);
e_scrollbar_resize(v->scrollbar.h, v->size.w - 12, 12);
/* I support dnd */
ecore_window_dnd_advertise(v->win.base);
ecore_window_show(v->win.main);
{
char *dir;
dir = v->dir;
v->dir = NULL;
e_view_set_dir(v, dir);
IF_FREE(dir);
}
if (!v->iconbar) v->iconbar = e_iconbar_new(v);
if (v->iconbar) e_iconbar_realize(v->iconbar);
v->changed = 1;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_update(E_View *v)
{
Evas_List l;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
if (v->changed)
{
for (l = v->icons; l; l = l->next)
{
E_Icon *icon;
icon = l->data;
}
if (v->drag.update)
{
ecore_window_move(v->drag.win, v->drag.x, v->drag.y);
v->drag.update = 0;
}
}
if (v->options.back_pixmap)
{
Imlib_Updates up;
up = evas_render_updates(v->evas);
/* special code to handle if we are double buffering to a pixmap */
/* and clear sections of the window if they got updated */
if (up)
{
Imlib_Updates u;
for (u = up; u; u = imlib_updates_get_next(u))
{
int x, y, w, h;
imlib_updates_get_coordinates(u, &x, &y, &w, &h);
ecore_window_clear_area(v->win.main, x, y, w, h);
}
imlib_updates_free(up);
}
}
else
evas_render(v->evas);
v->changed = 0;
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
static void
e_view_handle_fs(EfsdEvent *ev)
{
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
switch (ev->type)
{
case EFSD_EVENT_FILECHANGE:
switch (ev->efsd_filechange_event.changetype)
{
case EFSD_FILE_CREATED:
/* D("EFSD_FILE_CREATED: %i %s\n",
ev->efsd_filechange_event.id,
ev->efsd_filechange_event.file);
*/ e_view_file_added(ev->efsd_filechange_event.id,
ev->efsd_filechange_event.file);
break;
case EFSD_FILE_EXISTS:
/* D("EFSD_FILE_EXISTS: %i %s\n",
ev->efsd_filechange_event.id,
ev->efsd_filechange_event.file); */
e_view_file_added(ev->efsd_filechange_event.id,
ev->efsd_filechange_event.file);
break;
case EFSD_FILE_DELETED:
/* D("EFSD_FILE_DELETED: %i %s\n",
ev->efsd_filechange_event.id,
ev->efsd_filechange_event.file);
*/ e_view_file_deleted(ev->efsd_filechange_event.id,
ev->efsd_filechange_event.file);
break;
case EFSD_FILE_CHANGED:
/* D("EFSD_CHANGE_CHANGED: %i %s\n",
ev->efsd_filechange_event.id,
ev->efsd_filechange_event.file);
*/ e_view_file_changed(ev->efsd_filechange_event.id,
ev->efsd_filechange_event.file);
break;
case EFSD_FILE_MOVED:
/* D("EFSD_CHANGE_MOVED: %i %s\n",
ev->efsd_filechange_event.id,
ev->efsd_filechange_event.file);
*/ e_view_file_moved(ev->efsd_filechange_event.id,
ev->efsd_filechange_event.file);
break;
case EFSD_FILE_END_EXISTS:
{
E_View *v;
v = e_view_find_by_monitor_id(efsd_event_id(ev));
if (v) v->is_listing = 0;
/* D("EFSD_CHANGE_END_EXISTS: %i %s\n",
ev->efsd_filechange_event.id,
ev->efsd_filechange_event.file);
*/ }
break;
default:
break;
}
break;
case EFSD_EVENT_REPLY:
switch (ev->efsd_reply_event.command.type)
{
case EFSD_CMD_REMOVE:
break;
case EFSD_CMD_MOVE:
break;
case EFSD_CMD_SYMLINK:
break;
case EFSD_CMD_LISTDIR:
break;
case EFSD_CMD_MAKEDIR:
break;
case EFSD_CMD_CHMOD:
break;
case EFSD_CMD_GETFILETYPE:
/* D("Getmime event %i\n",
ev->efsd_reply_event.command.efsd_file_cmd.id); */
if (ev->efsd_reply_event.errorcode == 0)
{
E_Icon *ic;
E_View *v;
char *file;
file = NULL;
if ( (file = efsd_event_filename(ev)) )
{
file = e_file_get_file(file);
}
v = e_view_find_by_monitor_id(efsd_event_id(ev));
if ((v) && (file))
{
ic = e_icon_find_by_file(v, file);
if ((ic) &&
(ev->efsd_reply_event.data))
{
char *m, *p;
char mime[PATH_MAX], base[PATH_MAX];
m = ev->efsd_reply_event.data;
p = strchr(m, '/');
if (p)
{
strcpy(base, m);
strcpy(mime, p + 1);
p = strchr(base, '/');
*p = 0;
}
else
{
strcpy(base, m);
strcpy(mime, "unknown");
}
/* D("MIME: %s\n", m);
*/ e_icon_set_mime(ic, base, mime);
/* Try to update the GUI according to the file permissions.
It's just a try because we need to have the file's stat
info as well. --cK.
*/
e_icon_check_permissions(ic);
e_icon_initial_show(ic);
}
}
}
break;
case EFSD_CMD_STAT:
/* D("Stat event %i on %s\n",
efsd_reply_id(ev), efsd_reply_filename(ev)); */
/* When everything went okay and we can find a view and an icon,
set the file stat data for the icon. Then try to check the
permissions and possibly update the gui. It's just a try
because we need to have received the filetype info too. --cK.
*/
if (ev->efsd_reply_event.errorcode == 0)
{
E_Icon *ic;
E_View *v;
v = e_view_find_by_monitor_id(efsd_event_id(ev));
if (v)
{
ic = e_icon_find_by_file(v, e_file_get_file(efsd_event_filename(ev)));
if (ic)
{
ic->stat = *((struct stat*)efsd_event_data(ev));
e_icon_check_permissions(ic);
}
}
}
break;
case EFSD_CMD_READLINK:
2001-08-02 15:43:59 -07:00
if (ev->efsd_reply_event.errorcode == 0)
{
E_Icon *ic;
E_View *v;
char *file;
file = NULL;
if ( (file = efsd_event_filename(ev)) )
{
file = e_file_get_file(file);
}
v = e_view_find_by_monitor_id(efsd_event_id(ev));
if ((v) && (file))
{
ic = e_icon_find_by_file(v, file);
if ((ic) &&
(ev->efsd_reply_event.data))
e_icon_set_link(ic, (char*)efsd_event_data(ev));
e_icon_initial_show(ic);
}
}
break;
case EFSD_CMD_CLOSE:
break;
case EFSD_CMD_SETMETA:
break;
case EFSD_CMD_GETMETA:
/* D("Getmeta event %i\n",
efsd_reply_id(ev));*/
{
Evas_List l;
EfsdCmdId cmd;
cmd = efsd_event_id(ev);
for (l = views; l; l = l->next)
{
E_View *v;
int ok;
ok = 0;
v = l->data;
if (v->is_desktop) continue;
if (v->geom_get.x == cmd)
{
v->geom_get.x = 0;
if (efsd_metadata_get_type(ev) == EFSD_INT)
{
2001-08-02 15:43:59 -07:00
if (ev->efsd_reply_event.errorcode == 0)
{
if (efsd_metadata_get_int(ev,
&(v->location.x)))
{
ecore_window_move(v->win.base,
v->location.x,
v->location.y);
ecore_window_set_xy_hints(v->win.base,
v->location.x,
v->location.y);
}
}
}
ok = 1;
}
else if (v->geom_get.y == cmd)
{
v->geom_get.y = 0;
if (efsd_metadata_get_type(ev) == EFSD_INT)
{
2001-08-02 15:43:59 -07:00
if (ev->efsd_reply_event.errorcode == 0)
{
if (efsd_metadata_get_int(ev,
&(v->location.y)))
{
ecore_window_move(v->win.base,
v->location.x,
v->location.y);
ecore_window_set_xy_hints(v->win.base,
v->location.x,
v->location.y);
}
}
}
ok = 1;
}
else if (v->geom_get.w == cmd)
{
v->geom_get.w = 0;
if (efsd_metadata_get_type(ev) == EFSD_INT)
{
2001-08-02 15:43:59 -07:00
if (ev->efsd_reply_event.errorcode == 0)
{
if (efsd_metadata_get_int(ev,
&(v->size.w)))
{
ecore_window_resize(v->win.base,
v->size.w,
v->size.h);
v->size.force = 1;
}
}
}
ok = 1;
}
else if (v->geom_get.h == cmd)
{
v->geom_get.h = 0;
if (efsd_metadata_get_type(ev) == EFSD_INT)
{
2001-08-02 15:43:59 -07:00
if (ev->efsd_reply_event.errorcode == 0)
{
if (efsd_metadata_get_int(ev,
&(v->size.h)))
{
ecore_window_resize(v->win.base,
v->size.w,
v->size.h);
v->size.force = 1;
}
}
}
ok = 1;
}
else if (v->getbg == cmd)
{
v->getbg = 0;
if (efsd_metadata_get_type(ev) == EFSD_STRING)
{
if (ev->efsd_reply_event.errorcode == 0)
{
char buf[PATH_MAX];
IF_FREE(v->bg_file);
e_strdup(v->bg_file, efsd_metadata_get_str(ev));
sprintf(buf, "background_reload:%s", v->dir);
ecore_add_event_timer(buf, 0.5, e_view_bg_reload_timeout, 0, v);
}
}
}
if (ok)
{
if ((!v->geom_get.x) &&
(!v->geom_get.y) &&
(!v->geom_get.w) &&
(!v->geom_get.h) &&
(v->geom_get.busy))
{
E_Border *b;
v->geom_get.busy = 0;
if (v->bg)
e_background_set_size(v->bg, v->size.w, v->size.h);
if (v->options.back_pixmap) e_view_update(v);
b = e_border_adopt(v->win.base, 1);
b->client.internal = 1;
e_border_remove_click_grab(b);
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
}
}
break;
case EFSD_CMD_STARTMON_DIR:
/* D("Startmon event %i\n",
ev->efsd_reply_event.command.efsd_file_cmd.id);
*/ break;
case EFSD_CMD_STARTMON_FILE:
/* D("Startmon file event %i\n",
ev->efsd_reply_event.command.efsd_file_cmd.id);
*/ break;
case EFSD_CMD_STOPMON_DIR:
break;
case EFSD_CMD_STOPMON_FILE:
break;
default:
break;
}
break;
default:
break;
}
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
void
e_view_bg_load(E_View *v)
{
E_Background *bg;
char buf[PATH_MAX];
D_ENTER;
if (!v->prev_bg_file)
{
e_strdup(v->prev_bg_file, "/");
}
if (!v->bg_file)
{
e_strdup(v->bg_file, "");
}
else
{
/* relative path for bg_file ? */
if ((v->bg_file[0] != '/'))
{
sprintf(buf, "%s/%s", v->dir, v->bg_file);
FREE(v->bg_file);
e_strdup(v->bg_file, buf);
}
}
bg = e_background_load(v->bg_file);
if (!bg)
{
sprintf(buf, "%s/.e_background.bg.db", v->dir);
FREE(v->bg_file);
e_strdup(v->bg_file, buf);
bg = e_background_load(v->bg_file);
if (!bg)
{
if (v->is_desktop)
sprintf(buf, "%s/default.bg.db", e_config_get("backgrounds"));
else
sprintf(buf, "%s/view.bg.db", e_config_get("backgrounds"));
FREE(v->bg_file);
e_strdup(v->bg_file, buf);
bg = e_background_load(v->bg_file);
}
}
if (bg)
{
v->bg = bg;
v->bg_mod = e_file_mod_time(v->bg_file);
if (v->evas)
{
e_background_realize(v->bg, v->evas);
e_background_set_scroll(v->bg, v->scroll.x, v->scroll.y);
e_background_set_size(v->bg, v->size.w, v->size.h);
}
}
IF_FREE(v->prev_bg_file);
e_strdup(v->prev_bg_file, v->bg_file);
D_RETURN;
}
static void
e_view_bg_reload_timeout(int val, void *data)
{
E_View *v;
D_ENTER;
v = data;
if (!strcmp(v->prev_bg_file, v->bg_file))
{
time_t new_mod;
new_mod = e_file_mod_time(v->bg_file);
if (new_mod == v->bg_mod)
{
D("abort bg reload - same damn file\n");
D_RETURN;
}
}
if (v->bg)
{
int size;
e_object_unref(E_OBJECT(v->bg));
v->bg = NULL;
if (v->evas)
{
size = evas_get_image_cache(v->evas);
evas_set_image_cache(v->evas, 0);
evas_set_image_cache(v->evas, size);
}
e_db_flush();
}
e_view_bg_load(v);
D_RETURN;
UN(val);
}
void
e_view_bg_change(E_View *v, char *file)
{
char buf[PATH_MAX];
D_ENTER;
printf("change %s\n", file);
if (!(!strcmp(file, ".e_background.bg.db"))) return;
sprintf(buf, "background_reload:%s", v->dir);
ecore_add_event_timer(buf, 0.5, e_view_bg_reload_timeout, 0, v);
D_RETURN;
}
void
e_view_bg_add(E_View *v, char *file)
{
char buf[PATH_MAX];
D_ENTER;
printf("change %s\n", file);
if (!(!strcmp(file, ".e_background.bg.db"))) return;
sprintf(buf, "%s/%s", v->dir, file);
if (!strcmp(buf, v->bg_file)) D_RETURN;
IF_FREE(v->bg_file);
e_strdup(v->bg_file, "");
sprintf(buf, "background_reload:%s", v->dir);
ecore_add_event_timer(buf, 0.5, e_view_bg_reload_timeout, 0, v);
D_RETURN;
}
void
e_view_bg_del(E_View *v, char *file)
{
D_ENTER;
e_view_bg_change(v, file);
D_RETURN;
}
void
e_view_init(void)
{
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_ENTER;
ecore_event_filter_handler_add(ECORE_EVENT_MOUSE_DOWN, e_mouse_down);
ecore_event_filter_handler_add(ECORE_EVENT_MOUSE_UP, e_mouse_up);
ecore_event_filter_handler_add(ECORE_EVENT_MOUSE_MOVE, e_mouse_move);
ecore_event_filter_handler_add(ECORE_EVENT_MOUSE_IN, e_mouse_in);
ecore_event_filter_handler_add(ECORE_EVENT_MOUSE_OUT, e_mouse_out);
ecore_event_filter_handler_add(ECORE_EVENT_WINDOW_EXPOSE, e_window_expose);
ecore_event_filter_handler_add(ECORE_EVENT_KEY_DOWN, e_key_down);
ecore_event_filter_handler_add(ECORE_EVENT_KEY_UP, e_key_up);
ecore_event_filter_handler_add(ECORE_EVENT_MOUSE_WHEEL, e_wheel);
ecore_event_filter_handler_add(ECORE_EVENT_WINDOW_CONFIGURE, e_configure);
ecore_event_filter_handler_add(ECORE_EVENT_WINDOW_PROPERTY, e_property);
ecore_event_filter_handler_add(ECORE_EVENT_WINDOW_UNMAP, e_unmap);
ecore_event_filter_handler_add(ECORE_EVENT_WINDOW_VISIBILITY, e_visibility);
ecore_event_filter_handler_add(ECORE_EVENT_WINDOW_FOCUS_IN, e_focus_in);
ecore_event_filter_handler_add(ECORE_EVENT_WINDOW_FOCUS_OUT, e_focus_out);
ecore_event_filter_handler_add(ECORE_EVENT_WINDOW_DELETE, e_delete);
/* dnd source handlers */
ecore_event_filter_handler_add(ECORE_EVENT_DND_DROP_STATUS, e_dnd_status);
ecore_event_filter_handler_add(ECORE_EVENT_DND_DATA_REQUEST, e_dnd_data_request);
ecore_event_filter_handler_add(ECORE_EVENT_DND_DROP_END, e_dnd_drop_end);
/* dnd target handlers */
ecore_event_filter_handler_add(ECORE_EVENT_DND_DROP_POSITION, e_dnd_drop_position);
ecore_event_filter_handler_add(ECORE_EVENT_DND_DROP, e_dnd_drop);
ecore_event_filter_handler_add(ECORE_EVENT_DND_DROP_REQUEST, e_dnd_drop_request);
ecore_event_filter_idle_handler_add(e_idle, NULL);
e_fs_add_event_handler(e_view_handle_fs);
Alright, I spent some time now reading e17's code. Here's what I've changed, this is big, so read this carefully :) * I've added debugging macros for messages and function call tracing. Usage: D("Creating item %i %i %i\n", x, y, z); Define DEBUG to use the D macro. D_ENTER; D_RETURN; D_RETURN_(x); These are for call tracing. Use D_RETURN_(x) when returning something from a function. Define DEBUG_NEST to use this. * added iconbar header file to Makefile.am * added proper new()/cleanup() calls for E_Delayed_Action; * I've completely rewritten the object and observer handling. Bye bye macros, this was nasty. It'll be hard enough to avoid leaks with usecounting in C. We now basically have the same system as gtk. There's a clear separation of observer and object code now. An E_Object by itself has nothing to do with observing or being observed, therefore, there are now E_Observers and E_Observees that are derived from E_Object. IMPORTANT: The cleanup system now reflects the reference count system, therefore, all ..._free() calls are now static, because the destructor should never be called explicitly, but implicitly through e_object_unref(). The object handling now is as follows: - The cleanup functions clean up everything that is contained in a struct, but NOT the struct itself. Instead of the final free() call, they call the destructor of the base class. The calls will walk up the hierarchy and clean up what's contained in every struct, and the final e_object_cleanup() will free the structure itself. E_Delayed_Action is a good example. - The only calls that influence the reference count are e_object_ref() and e_object_unref(). If you need to do things before an object gets destroyed, you can query the use count using e_object_get_usecount() and check if it's equal to 1. So this: OBJ_UNREF(b); OBJ_IF_FREE(b) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); OBJ_FREE(b); } now is this: if (e_object_get_usecount(E_OBJECT(b)) == 1) { ecore_window_reparent(e->win, 0, 0, 0); e_icccm_release(e->win); } e_object_unref(E_OBJECT(b)); object.h and observer.h are completely commented, it shouldn't be too hard to understand. This'll need to be documented in the manual anyway. * E_Objects are now used in lots of places where void* were used as pointers to objects before, especially in the actions code. This is obviously better, as it will generate compiler warnings when people want to pass things to functions that expect E_Objects. This could probably be more restrictive. * Added typedefs for the function prototypes in E_Action_Impl. Those fat signatures were just painful to read in the function declarations/implementations. * I've also tried to give parameters more useful names. Calling an object "o" is a lot of fun when you want to grep for it. * Included is also Graham's latest menu.c patch. Sorry for the delay, Graham. * I've added checks to the menu code that make sure that menus don't pop up when they're empty (which resulted in a little useless rectangle). I guess that's it for now. Sorry if I broke anything, but this was necessary imho. SVN revision: 5605
2001-11-02 09:07:52 -08:00
D_RETURN;
}
/*
* send the dnd data to the target app
*
* URL formatting per RFC 1738
*/
static void
e_dnd_data_request(Ecore_Event * ev)
{
Ecore_Event_Dnd_Data_Request *e;
/*
* typedef struct _ecore_event_dnd_data_request
* {
* Window win, root, source_win;
* int plain_text;
* Atom destination_atom;
* } Ecore_Event_Dnd_Data_Request;
*/
Evas_List l;
char hostname[PATH_MAX];
D_ENTER;
/* Need hostname for URL (file://hostname/...) */
if(gethostname( hostname, PATH_MAX))
{
/* failed... Default to 'localhost' */
strcpy( hostname, "localhost");
}
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
Evas_List ll;
char *data = NULL;
int size = 3;
int idx = 0;
/* Me, my null, and an extra for the end '/r/n'... */
data = NEW(char, size);
*data = 0;
v = l->data;
if (e->win == v->win.base)
{
for (ll = v->icons; ll; ll = ll->next)
{
E_Icon *ic;
ic = ll->data;
if (ic->state.selected)
{
int ic_size;
/* Size = 'file://' + 3 strings + host delimiter '/' and '\r\n' end. */
ic_size = 7 + strlen(hostname) + strlen(v->dir) + strlen(ic->file)+3;
size += ic_size;
REALLOC(data, char, size);
sprintf( data+idx, "file://%s%s/%s\r\n", hostname, v->dir, ic->file);
idx += ic_size;
}
}
if(v->drag.drop_mode == E_DND_COPY)
ecore_dnd_set_mode_copy();
else
ecore_dnd_set_mode_move();
ecore_dnd_set_data(e->win);
ecore_dnd_send_data(
e->source_win, e->win,
data, size,
e->destination_atom,
/* uri-list, not plain-text */
0
);
}
FREE(data);
}
D_RETURN;
}
static void
e_dnd_drop_end(Ecore_Event * ev)
{
Ecore_Event_Dnd_Drop_End *e;
/*
* typedef struct _ecore_event_dnd_drop_end
* {
* Window win, root, source_win;
* } Ecore_Event_Dnd_Drop_End;
*/
Evas_List l;
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.base)
{
ecore_window_dnd_finished();
e_dnd_drop_request_free();
}
}
D_RETURN;
}
static void
e_dnd_drop_position(Ecore_Event * ev)
{
Ecore_Event_Dnd_Drop_Position *e;
/*
* typedef struct _ecore_event_dnd_drop_position
* {
* Window win, root, source_win;
* int x, y;
* } Ecore_Event_Dnd_Drop_Position;
*/
Evas_List l;
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.base)
{
if( e->win != e->source_win )
{
/* send XdndStatus */
ecore_window_dnd_send_status_ok(v->win.base, e->source_win,
v->location.x, v->location.y,
v->size.w, v->size.h
);
}
/* todo - cache window extents, don't send again within these extents. */
}
}
D_RETURN;
}
static void
e_dnd_drop(Ecore_Event * ev)
{
Ecore_Event_Dnd_Drop *e;
/*
* typedef struct _ecore_event_dnd_drop
* {
* Window win, root, source_win;
* } Ecore_Event_Dnd_Drop;
*/
Evas_List l;
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.base)
{
/* Dropped! Handle data */
e_dnd_handle_drop( v, dnd_pending_mode );
ecore_window_dnd_finished();
e_dnd_drop_request_free();
}
}
D_RETURN;
}
static void
e_dnd_drop_request(Ecore_Event * ev)
{
Ecore_Event_Dnd_Drop_Request *e;
/*
* typedef struct _ecore_event_dnd_drop_request
* {
* Window win, root, source_win;
* int num_files;
* char **files;
* int copy, link, move;
* } Ecore_Event_Dnd_Drop_Request;
*/
Evas_List l;
D_ENTER;
e = ev->event;
for (l = views; l; l = l->next)
{
E_View *v;
v = l->data;
if (e->win == v->win.base)
{
/* if it exists, we already have the data... */
if( !dnd_files )
{
int i;
dnd_files = NEW_PTR(e->num_files);
/* copy the file list locally, for use in a dnd_drop */
for( i=0; i < e->num_files; i++ )
dnd_files[i] = strdup( e->files[i] );
dnd_num_files = e->num_files;
if( e->copy )
dnd_pending_mode = E_DND_COPY;
else if( e->move )
dnd_pending_mode = E_DND_MOVE;
else if( e->link )
dnd_pending_mode = E_DND_LINK;
else
dnd_pending_mode = E_DND_ASK;
}
/*
printf( "drop-req %d-[c%dm%dl%d]--%s--\n", e->num_files,
e->copy, e->move, e->link,
e->num_files ? e->files[0] : "None"
);
*/
}
}
D_RETURN;
}
static void
e_dnd_drop_request_free(void)
{
if (dnd_files)
{
int i;
for (i = 0; i < dnd_num_files; i++)
FREE(dnd_files[i]);
FREE(dnd_files);
dnd_num_files = 0;
}
}
static void
e_dnd_handle_drop( E_View *v, E_dnd_enum dnd_pending_mode )
{
char hostname[PATH_MAX];
int in, out;
char *filename;
D_ENTER;
/* Need hostname for URL (file://hostname/...) */
if(gethostname( hostname, PATH_MAX))
{
/* failed... Default to 'localhost' */
strcpy( hostname, "localhost");
}
/* Make space for destination in file list */
dnd_num_files++;
REALLOC_PTR(dnd_files, dnd_num_files);
dnd_files[dnd_num_files-1] = NULL;
/* Verify files are local, convert to non-URL */
for( in=0, out=0; in<dnd_num_files-1; in++ )
{
filename = e_util_de_url_and_verify( dnd_files[in], hostname );
/* Need a overlap safe copy here, like memmove() */
if( filename )
memmove( dnd_files[out++], filename, strlen(filename)+1 );
}
/* Append destination for efsd */
if( dnd_files[out] )
FREE( dnd_files[out] );
dnd_files[out++] = strdup( v->dir );
switch( dnd_pending_mode )
{
case E_DND_COPY:
/* Copy files */
efsd_copy( e_fs_get_connection(), out, dnd_files,
efsd_ops(0) );
break;
case E_DND_MOVE:
efsd_move( e_fs_get_connection(), out, dnd_files,
efsd_ops(0) );
break;
default:
/* nothing yet */
break;
}
D_RETURN;
}