I.e walk using a ptr instead of arr[i] all the time. This should make
things faster. Also as part of the change, fixed char_index to be
an offset inside the string object, and not a general offset which was
bad.
SVN revision: 57241
process calls hapen and leave some objects uncalculated, but
removed from the array and marked as needing a calc thus never
getting back into the array.
SVN revision: 57234
NOTE: eet_sync just force the write of the eet file to the disk.
That's exactly what eet_close will do also. eet_sync is just to
be used when you are planning to still use it and don't want to
close it soon.
SVN revision: 57233
Subject: [E-devel] [PATCH] ecore_imf_context_canvas_get,
ecore_imf_context_window_get
This patch includes the code to add ecore_imf_context_canvas_get and
ecore_imf_context_window_get.
These APIs are coupled with the ecore_imf_context_canvas_set and
ecore_imf_context_window_set.
SVN revision: 57203
Subject: RE: [E-devel] One bug in _edje_key_down_cb func (in
edje_entry.c)
I found one bug in _edje_key_down_cb func (in edje_entry.c)
When I input "Backspace" in an entry object, _signal_cursor_chagned_func(in
elm_entry.c) is not called.
This is because the following code in _edje_key_down_cb func (in edje_entry.c)
is not called.
if ((evas_textblock_cursor_compare(tc, en->cursor)) && (!cursor_changed))
_edje_emit(ed, "cursor,changed", rp->part->name);
Why not called ? Because ~ evas_textblock_cursor_compare(tc, en->cursor)
returns 0.
(i.e. old cursor and current cursor have the same position information)
I think old cursor should keep its position without any updating.
But in evas_textblock_cursor_char_delete() func, called by _backsapce func
(in edje_entry.c) -
old cursor's postion is updated as the current position.
SVN revision: 57199
if Ecore_X_Image-->XImage does not exist, we call
_ecore_x_image_shm_create to create it via shm, BUT that function
can return a NULL XImage if shm is not supported, so we need to check
the return of that, else we are calling XShmPutImage with no XImage.
(NB: This should probably be backported to 1.0...if someone could
handle that please ?)
SVN revision: 57130
Added standalone continuation bytes tests.
Added isolated starting sequences tests
Added incomplete sequences tests
Added Overlong representations tests.
And I think that's it. Still need to add tests for surrogate pairs.
SVN revision: 57123