Summary:
This patch fixes a memory leak in _elm_config_profile_name_get() where
the data returned from eet_read is not being free'd, causing valgrind
to spit out:
```
==11610== 9 bytes in 1 blocks are definitely lost in loss record 16 of
572
==11610== at 0x483777F: malloc (vg_replace_malloc.c:299)
==11610== by 0x4D9DE5B: eina_strbuf_common_string_steal
(eina_strbuf_common.c:828)
==11610== by 0x4D5BC31: eina_binbuf_string_steal
(eina_binbuf_template_c.x:187)
==11610== by 0x58646A1: eet_read_cipher (eet_lib.c:1991)
==11610== by 0x5864775: eet_read (eet_lib.c:2022)
==11610== by 0x49CD4D2: _elm_config_profile_name_get
(elm_config.c:2207)
==11610== by 0x49CD5F8: _elm_config_profile_save (elm_config.c:2237)
==11610== by 0x49D2F6E: _elm_config_sub_init (elm_config.c:4467)
==11610== by 0x4AC0B7D: elm_quicklaunch_sub_init (elm_main.c:840)
==11610== by 0x4ABFE3A: elm_init (elm_main.c:409)
==11610== by 0x138143: main (test.c:1379)
```
@fix
Reviewers: raster, cedric, ManMower
Reviewed By: raster
Subscribers: #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7646
this avoid warings about keys not beeing used in the fallback function.
Reviewed-by: Daniel Kolesa <daniel@octaforge.org>
Differential Revision: https://phab.enlightenment.org/D7630
It appears that NULL is not really usefull here, as this will lead to
errors later on. (void) will force the free function to just do nothing
and be later converted to a NOP.
Reviewed-by: Daniel Kolesa <daniel@octaforge.org>
Differential Revision: https://phab.enlightenment.org/D7629
this removes the need for the calling a Init function.
Reviewed-by: Felipe Magno de Almeida <felipe@expertisesolutions.com.br>
Differential Revision: https://phab.enlightenment.org/D7556
Summary:
until to today you had to call init functions and a run function which
were static function in a class called Efl.Ui.Config.
However, calling those init functions there is not really OOP style.
Right now things have changed into a manner where you are defining you
application class with inheriting from the Application /
SimpleApplication abstract.
This enables you to call launch() on your application class, calling
launch there leads to a call to the args function, you can call and use
the Efl classes in there, everything is booted up.
Option parsing and dependency start can still be done in the main method
or application constructor, just ensure that you never call any efl
class / function outside the launch function.
A commit that demonstrates the usage can be found at
ref T7204
https://git.enlightenment.org/tools/examples.git/log/?h=devs/bu5hm4n/POC
Reviewers: felipealmeida, segfaultxavi, Jaehyun_Cho, cedric
Reviewed By: segfaultxavi
Subscribers: zmike, woohyun, akanad, lauromoura, #reviewers, #committers
Tags: #efl_language_bindings
Maniphest Tasks: T7204
Differential Revision: https://phab.enlightenment.org/D7495
Summary:
Since glib 2.58 g_type_class_add_private is deprecated and this throws
an ugly warning in our build. Not being a glib expert I based this on
patch in freeverb which solves the same issue for them:
90625953f2
Videos in rage are still playing fine for me, but double checking during
review would be appreciated.
Reviewers: raster
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7638
Summary:
If widget use both hint_align(HINT_FILL) and hint_max together, there is no way
to set to hint_align. See, elementary_test -to 'efl.ui.box' -
"Button with a quite long text." button. you can control button position
using "Box align" slider, but this is not proper implementation. When there are
two widget which have hint_align(HINT_FILL) and hint_max, those positions are
determined by box_align rather than hint_align. it means widget align cannot be
set individually.
To solve this problem, this patch add hint_fill property. in order to avoid
conflict with legacy API named evas_object_size_hint_fill_set(), it only works
for EO widgets(made by efl_add).
Also, EFL_GFX_SIZE_HINT_FILL is removed.
@feature
Test Plan:
elementary_test -to 'efl.ui.box'
elementary_test -to 'efl.ui.table'
Reviewers: jpeg, Hermet, Jaehyun_Cho, raster, barbieri
Reviewed By: Jaehyun_Cho
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Maniphest Tasks: T3912
Differential Revision: https://phab.enlightenment.org/D7409
Summary: The trailings end up in the final version, which causes it to create a invalid XML file.
Reviewers: bu5hm4n, woohyun, segfaultxavi
Reviewed By: bu5hm4n
Subscribers: cedric, segfaultxavi, woohyun, #reviewers, bu5hm4n, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7613
This reverts commit 6c4e49d970.
Hell no - this removes loading of system config profile. the entire
premise of the patch is wrong. it isn't reading the code at all... how
this got through review is rather amazing...
_set_a_without_b() makes a list clone internally, so cloning the list
first will leak a copy.
Signed-off-by: Derek Foreman <derek.foreman.samsung@gmail.com>
Reviewed-by: Marcel Hollerbach <marcel-hollerbach@t-online.de>
Differential Revision: https://phab.enlightenment.org/D7622
I guess this used to do something...
Now it's a list that we never add anything to but frequently try to remove
things from.
Signed-off-by: Derek Foreman <derek.foreman.samsung@gmail.com>
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Reviewed-by: Chris Michael <cp.michael@samsung.com>
Differential Revision: https://phab.enlightenment.org/D7621
Not that these deletes ever walk long lists, but it's trivial to stop
doing it entirely.
Signed-off-by: Derek Foreman <derek.foreman.samsung@gmail.com>
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Reviewed-by: Chris Michael <cp.michael@samsung.com>
Differential Revision: https://phab.enlightenment.org/D7611
They're only ever on a single list, and never counted. inlist makes
more sense.
Signed-off-by: Derek Foreman <derek.foreman.samsung@gmail.com>
Reviewed-by: Chris Michael <cp.michael@samsung.com>
Differential Revision: https://phab.enlightenment.org/D7610
We know the item is in there, so we can remove it directly.
Signed-off-by: Derek Foreman <derek.foreman.samsung@gmail.com>
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D7609
it seems that this class does not have a meaning when created just like
this. Other classes using it are even abstract, which means, this class
should also be abstract. This is done in order to support that a
abstract class should only contain abstract
ref T7240
Reviewed-by: Xavi Artigas <xavierartigas@yahoo.es>
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D7601
the efl_canvas_group should be abstract in order to support other
classes which are inheriting from it, which are abstract. This is done
in order to support only abstract parents from abstract classes.
ref T7240
Reviewed-by: Xavi Artigas <xavierartigas@yahoo.es>
Differential Revision: https://phab.enlightenment.org/D7600
There is no point in having the object instancable. However, in order to
support bindings, we need to ensure that every abstract does have only
abstracts as inherit-parents, thus making this class abstract.
ref T7240
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D7599
Summary:
Previously, any unhandled Eina_Error would cause an exception
to be thrown when the control returned to C#.
This commit changes this behavior to only raise it when an exception
went unhandled from a C# callback back to C, like in an event handler,
for example.
Test Plan: run tests
Reviewers: segfaultxavi, Jaehyun_Cho, felipealmeida
Reviewed By: Jaehyun_Cho
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7537
Summary:
Previously, we just converted through PtrToStructure, which didn't have
the full marshalling info converting from the internal one to the
external.
This fixes the usage of the Efl.Loop.arguments event.
Also renamed the ToExternal methods to ToManaged, to make clearer that
the output struct is the one intended to be used from the managed code.
Also fixed a minor styling in the generated code (making it easier to be
inspected).
Depends on D7538
Reviewers: segfaultxavi, felipealmeida, bu5hm4n
Reviewed By: bu5hm4n
Subscribers: bu5hm4n, cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7539
Summary:
"type" in .eo is converted to "struct" in .eo.cs.
Since the type name in .eo is the same with the struct name .eo.cs,
'_' is removed from the converted struct in .eo.cs for C# naming
convention.
For example, Efl.Callback_Priority is defined in efl_object.eo and
the name is converted to Efl.CallbackPriority in efl_object.eo.cs.
Efl.Access.StateSet in workaround.cs causes duplicated definition
with this patch so Efl.Access.StateSet in workaround.cs is removed.
Test Plan: Compile with autogen.sh --enable-csharp-bindings
Reviewers: lauromoura, segfaultxavi, felipealmeida
Reviewed By: felipealmeida
Subscribers: segfaultxavi, cedric, #reviewers, #committers, woohyun
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7597
so drop taregts liek rage stopped working - something was wrong with
type handling and fetching selections at all and so on... i've kicked
it back into working again. it's not perfect but it's improved. at
least simple thgins like being a dnd target for dnd from a filemanager
work again which is probably the biggest use case.
@fix
Summary: Without this a new overlay content aren't displayed into the map widget.
Reviewers: zmike, devilhorns
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7457
If EFL is built with tslib support, which is used by ecore_fb, then we
need to add a dependency on tslib else the build will fail to compile
(undefined reference to ts_open function).
Summary:
new Eina.Value(0) is a special case. The 0 is silently converted
to an enum (Eina.ValueType) and therefore the call is ambiguous
with the 0 being first converted to an Eina.Value via the implicit
conversion operator (calling the Eina.Value deep copy constructor).
Adding constructors for all supported types solves the problem because
they have higher priority. Also, they avoid one deep copy of the
Eina.Value.
Includes test case to catch this problem in the future. This was discovered
in the tutorials, where new Eina.Value(0) is being used.
Test Plan:
The src/efl_reference_core_event.exe example from the examples repo was
not compiling before, and now it is.
make check and make examples still work as expected.
Reviewers: lauromoura
Reviewed By: lauromoura
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D7598
This introduces a new keyword called required. It only works on mixins.
You can specify a list of regular/abstract classes in there.
Classes specified after the required keyword are later used to verify
the usage of the mixin. With this feature a mixin can define a list of
types that the inheriting object (the object that inherits from a mixin)
needs to fullfill, if one class that is required is not in the
implemented classes, then eolian will bail out.
Differential Revision: https://phab.enlightenment.org/D7584
if clipped color changed and not object color on textblock objects, then
the updates could be missed. This bug seems to have been here a while
unnoticed.
This patch refers the patch 4e044fd9c2.
Even if a object is deleted, its render_pre function could be called.
Especially, Evas Textblock is calling relayout() function if there are
any changes. In relayout() function, it creates at least one paragraph
and one line. Normally, all of paragraphs and its line should be removed
when Evas Textblock object is deleted. But, paragraphs and lines which are
created after deleting its object never be free'd.
@fix
Differential Revision: https://phab.enlightenment.org/D7427
The evas_norender updates the canvas internal objects.
But efl_ui_win does not evaluate its internal objects, when evas_norender is
called before showing, after resizing as below.
evas_object_resize(win, 300, 600);
evas_norender(evas_object_evas_get(win));
evas_object_show(win);
This problem could be verified by checking if a resize function of internal
object is called or not.
minw,h is 0 in _elm_win_resize_objects_eval but deferred_resize_job is TRUE.
evas_norender -> _window_layout_stack -> _elm_win_resize_objects_eval
So if _elm_win_resize_objects_eval does not return if deferred_resize_job is
TRUE even if minw,h is 0, and calls _elm_win_resize_job, then it will work.
_elm_win_resize_objects_eval -> _elm_win_resize_job ->
evas_object_geometry_set -> _efl_canvas_group_group_need_recalculate_set ->
_window_layout_stack -> evas_object_geometry_set -> resize function.
I have checked this behavior without elementary. It seems that evas_norender
works between resize and show in this case. Let me share examples.
ecore_evas_resize(ee, 100, 100);
evas_norender(evas);
ecore_evas_show(ee);
Differential Revision: https://phab.enlightenment.org/D7425
There is a option(--with-elementary-base-dir) on autotools
this patch provides the exactly same thing that autotools does on meson
Differential Revision: https://phab.enlightenment.org/D7580
My use case here is having elementary_test being loaded by exactness.
That has been working fine with our autotools based build system but no
longer with meson.
Bringing fPIC and pie back as options allows exactness to dlopen the
binary and work again.
Differential Revision: https://phab.enlightenment.org/D7587
Signed-off-by: Stefan Schmidt <s.schmidt@samsung.com>