Summary:
If there is an empty (unused) <defs /> inside the svg file, this can cause problems.
<defs> node is managed separately in loader->def.
So it doesn't have to be added to loader's list.
Test Plan: N/A
Reviewers: Hermet, smohanty
Reviewed By: Hermet
Subscribers: cedric, #reviewers, #committers, kimcinoo, herb
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12127
coverity was complaining that the DIRECTION ACCESS macro might access
over a negative direction. However, complement of a 2D direction is
always a 2D direction. But coverity cannot detect that.
Summary:
mo_path can using for fprintf in using_file function without initialize.
fix coverity.
Reviewers: Hermet, raster
Reviewed By: Hermet
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12120
the xetension list for video files was far too small. it menas it
missed many kinds of video files. this fixes that to have a much wider
range/list.
@fix
Summary:
The style padding was included in native width(not in native height)
and formatted height(not in formatted width).
This is so weired. In addition, there is no enough document about
the relation between formatted size, native size and the style padding.
This issue is caused by a confusing code which is about how to handle
the style padding on item's width and height.("x_adjustment"!)
When Evas calculates "c->wmax" in line finalization stage, it explicitly subtract
style padding from line width. So, I assumed the formatted size has not to include
style padding. It is same for the native size.
The style padding will not be included in formatted size and native size by this commit.
@fix
Test Plan:
A test case is included in this commit.
Evas_Object *tb = evas_object_textblock_add(evas);
newst = evas_textblock_style_new();
evas_textblock_style_set(newst, "DEFAULT='font=Sans font_size=50 color=#000 text_class=entry'");
evas_object_textblock_style_set(tb, newst);
evas_object_textblock_text_markup_set(tb, "<style=far_soft_shadow>Test</>");
evas_object_textblock_style_insets_get(tb, &l, &r, &t, &b);
fail_if((l != 0) || (r != 4) || (t != 0) || (b != 4));
/* Size with style padding */
evas_object_textblock_size_formatted_get(tb, &w, &h);
evas_object_textblock_size_native_get(tb, &nw, &nh);
/* It is non-sense if the following condition is true. */
fail_if((w + l + r == nw) && (h == nh + t + b));
Reviewers: raster, ali.alzyod, woohyun, bowonryu
Reviewed By: ali.alzyod
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12110
Summary:
if we have rainbow flag emoji (🏳️‍🌈)
we can use mouse/keyboard to move cursor inside it because we break it into two clusters, we break on 1F308,
This is wrong as we should treat emoji as a single cluster (based on rules mentioned in Unicode segmentation standard “Do not break within emoji modifier sequences or emoji ZWJ sequences” (https://unicode.org/reports/tr29/#GB11 )).
this issue happens because we don’t give 1F308 its correct grapheme break property value, I think this is a bug in the unibreak library as this Unicode 1F308 should have word break class value equals to Glue_After_ZWJ (based on https://www.unicode.org/reports/tr29/tr29-31.html#Glue_After_Zwj_WB and http://unicode.org/Public/emoji/5.0/emoji-zwj-sequences.txt) which will not make it break and we will get a single cluster.
I noticed that the current unibreak lib used in EFL seems to implement Unicode 9 (latest is Unicode 13) which uses obsolete and unused grapheme break property, such as E_Modifier & Glue_After_ZWJ, so if a new emoji introduced (rainbow flag was introduced after Unicode 9) and based on Unicode 9 it should use property E_Modifier or Glue_After_ZWJ we will have issue with it.
So I have updated unibreak lib using latest released version of unibreak (4.2) which implement Unicode 12.
I needed to remove **BREAK_AFTER(i)** to pass the tests in D1140 as spaces do not break on latest update (also related to T995).
{F3868712}
this should fix T8665 & T8688
Reviewers: ali.alzyod, woohyun, bowonryu, zmike, segfaultxavi, bu5hm4n
Reviewed By: ali.alzyod
Subscribers: segfaultxavi, cedric, #reviewers, #committers
Tags: #efl
Maniphest Tasks: T8665
Differential Revision: https://phab.enlightenment.org/D11743
the test on the .bat file was too early
Reviewed-by: João Paulo Taylor Ienczak Zanette <joao.tiz@expertisesolutions.com.br>
Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D12122
Add gettimeofday implementation based on the mingw one, using a more acurate function if Windows >= 8
Reviewed-by: João Paulo Taylor Ienczak Zanette <joao.tiz@expertisesolutions.com.br>
Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D12111
Implement directly monotonic clock on Windows, to have more accurate results
Reviewed-by: João Paulo Taylor Ienczak Zanette <joao.tiz@expertisesolutions.com.br>
Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D12112
Although the [remove manpage](https://linux.die.net/man/3/remove) states that `remove(...)` deletes
either a file or a directory, this is not true in Windows as it can be seen in
[MSDN docs for
remove](https://docs.microsoft.com/en-us/cpp/c-runtime-library/reference/remove-wremove?view=vs-2019):
> **(Function description)**
>
> Delete a file.
>
> **Return Value**
>
> Each of these functions returns 0 if the file is successfully deleted.
> Otherwise, **it returns -1 and sets errno either to EACCES to indicate that the
> path** specifies a read-only file, //**specifies a directory**//, or the file
> is open, or to ENOENT to indicate that the filename or path was not found.
This implementation detail caused the Eina test to fail and not removing the
temporary directory.
This patch changes the use of `remove` to the directory-specific `rmdir`, which
is guaranteed to remove the directory. Additionally, it also deletes the
Eina_TmpStr that holds the temporary directory path.
Reviewed-by: Vincent Torri <vincent.torri@gmail.com>
Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D12115
119d9f39dd fixed async loads when head
skip was on but it broke informing the evas image object of the file
handles etc. by losing the skip head flag in the image entry. this
fixes that
changed files were not being detected as stas was being skipped if
skip head was enabled. this means we totally didnt see changes to
files if loaded this way until caches were cycled/flushed/removed
@fix
it wasn't being enabled thus missing an entire optimization limiting
stats to "once per frame" etc. to avoid overly syscall+io happy code
from having as much of an impact
@fix
it seems some libc's or systes dont reset signal mask blocks on
exec(). this unblocks manuaklly jus befor eexec, moves blocking to
before fork and adds a "nuke all signal handlers" in the child process.
fixes T8797
Summary:
It is able to get orientation information of inlined image object.
This information should be same during its life time.
The inlined image object got correct information only after size
calculation, so there is a kind of timing issue.
An example will be attached for more details.
I am not sure what the regression bug mentioned on D6855, but
this will keep compatibility of behavior.
Actually it seems that the compatibility was broken by D9686
which does not call _efl_ui_image_sizing_eval immediately.
Test Plan: {F3947703}
Reviewers: Hermet, jsuya, herb
Reviewed By: Hermet
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12114
Summary:
This patch fixes the issue that any application is not rescaled
when font or font size is changed by elementary_config.
Test Plan:
1. run elementary_test
2. run elementary_config
3. change the font in elementary_config
4. check the button size in elementary_test
Reviewers: id213sin, ali.alzyod, devilhorns, raster
Reviewed By: devilhorns, raster
Subscribers: devilhorns, cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12117
Sometimes display ref count is not zero after wl_display_dispatch_pending.
For example, wl_data_source_listener's callback is called by wl_display_dispatch_pending.
Display ref count is not zero but display is cleaned up, it makes double free corruption.
Reviewed-by: Christopher Michael <devilhorns@comcast.net>
Differential Revision: https://phab.enlightenment.org/D12109
in theory signal handlers could kick in after fork and before exec...
so block them until we're exec'd or exited so they don't change any
program state.
this won't happen as the setup should not have the # of fd's found in
2st pass and 2nd pass for a dirfd change as nothing should interfere.
comment to that effect and add a paranoid change num_closes to the
number found
post flush doesnt cleanr evas->rendering until after it calls post
flush callbacks. any post flush callback that neexts so access data
and block waiting for an async render tyo be done thus blocks forever.
this adds a 0.2 sec timeout in case so we at last march on with
hiccups instead of totally stalling AND clears evas->rendering before
calling the callbacks.
@fix
while it requires an extra wakeup from vsync thread -> e and this adds
a bit of jitter, it's mroe reliable in the face of "missed a frame"
timings and doesnt degrade as badly.
At least on Windows (didn't have the time to test on Linux yet),
running tests, even if they passed, there would be an Eina error on logs
pointing that a temporary file wasn't closed:
```
ERR:eina_file ../src/lib/eina/eina_file_common.c:1137 eina_file_shutdown() File [C:/Users/joao_/AppData/Local/Temp/aaaa_file_test_EBpVea] still open 1 times !
```
In the end, it was the `eina_file_test_unlink` that would create a temporary
file but never close it, being caught only by `eina_shutdown()`.
Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Reviewed-by: Vincent Torri <vincent.torri@gmail.com>
Differential Revision: https://phab.enlightenment.org/D12062
Reviewed-by: João Paulo Taylor Ienczak Zanette <joao.tiz@expertisesolutions.com.br>
Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D12108
In some systems (such as Windows), OpenSSL raises an error about "No Applink"
(see ["I've compiled a program under Windows and it crashes: why?" in OpenSSL
FAQ](https://www.openssl.org/docs/faq.html#PROG3)). Including only
openssl/applink.c didn't work, so the solution was to replace `FILE*`
interfaces with OpenSSL's BIO API which also contains file operations.
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D12103
Summary: Resolve rendering e vowel (0x1031) with Myanmar(Burmese) with zero width non joiner (0x200C)
Test Plan: ninja test
Reviewers: woohyun, bowonryu
Reviewed By: bowonryu
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12102
Summary:
If shadow format changed using style_apply , the event EFL_CANVAS_TEXTBLOCK_EVENT_CHANGED will not be fired.
also added tests for it.
Reviewers: ali.alzyod, woohyun
Reviewed By: ali.alzyod
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D11078
I found wrong API group name in elm_win and elm_glview and fixed them.
Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D12075
if this is set to 0 the next iteration in the upper frame event would
decrement the 0 again leading to a overflow making the iteration and
callback array overflow.
Long story short: set ifx to 1 to prevent overflow, test added.
fixes T8787
Differential Revision: https://phab.enlightenment.org/D12101
If we don't have a buffer here you CAN lose the race when scrolling.
As there is a check later in the flow, we can just use a padding
of 64 instead of checking the range. This ensures that scrolling is
smooth and content is always rendered throughout an aggressive
scroll.
Summary:
Current version checking is done with GetVersion, which is now deprecated and issues the warning:
```
'GetVersion' is deprecated [-Wdeprecated-declarations]
```
The first approach of this commit was to replace GetVersion call with the newer (and more reliable) `IsWindowsVistaOrGreater()` from versionhelpers.h, but nowadays it makes more sense to actually not even make that check, since it is more than unlinkely to happen that someone tries to compile/run EFL in Windows XP.
Reviewers: vtorri, stefan_schmidt, raster
Reviewed By: vtorri, raster
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12095
Summary:
By default a sub-suface is synchronized mode.
So when a sub-surface is created, its sync value should be TRUE.
If the E works as specification, user cannot make it work as
desynchronized mode by calling subsurface.set_sync(FALSE).
[Reference]
https://github.com/wayland-project/wayland/blob/master/protocol/wayland.xml
is telling "A sub-surface is initially in the synchronized mode."
signed-off-by: Shawn Lee <shiin.lee@samsung.com>
Reviewers: Hermet, zmike, devilhorns, raster
Reviewed By: devilhorns
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12098
we have these special counts which are > 0 when there is a callback
subscribed. THey are currently evalulated in _callback_call. However,
we can also skip the entire call from inside eo_base_class as we are
having the fields there as well.
This way we are skipping the obj pointer lookup and vtable lookup.
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D12079
a little overfiew to how eos event emission is working:
- In eo there are 2 types of events restart events and normal events.
Normals are *always* emitted from callback_max to 0
Restarts are emitted from the previous emissions current idx to 0. The
previous emission will stop then
- To keep track of the previous event emissions for restarts there was
an eina inlist.
- To keep track of other things in eo, there was an additional eo event
stack frame (newly inserted events etc. etc.)
This commit now uses this event stack frame for implementing the details
about the restart events. This has the advatage that every
efl_object_data contains one byte less, and the up to date keeping of
the id's in the restart infrastructure is not needed anymore, making the
whole event emission code less instructions and faster.
How this now works:
- Every emission has a event emission frame, when the emission starts,
the frame is added, next points to the next older event emission. In
this event stack frame we are storing the current idx, the description
and a few other things that are not really relevant at this point.
- when a restart event is started, this event stack is searched from the
current frame up to the next one featuring the same event description.
The event frame stack is then remembered, if the event emission is
done, the current idx (normally 0 or something higher when callback
was stopped) is then copyied back to the earlier frame.
- Based on the nature of the event frame stacks, the idx is updated
every iteration. And for the restart events, we only need to know the
next, which removes the need of permanently updating the idx in the
current stack.
- The event frame stack is not allocated on the heap, hence more things
in there does not matter at all.
Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D12097
this is a patch to fix a potentional error by null dereferencing.
Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D12087
this is a patch to fix a potentional error by null dereferencing.
Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D12093
Summary: this is a patch to fix a potentional error by null dereferencing.
Reviewers: jsuya, bu5hm4n, Hermet
Reviewed By: jsuya
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12086
Summary: this is a patch to fix a potentional error by null dereferencing.
Reviewers: jsuya, bu5hm4n, cedric, Hermet
Reviewed By: jsuya
Subscribers: cedric, #reviewers, #committers
Tags: #efl
Differential Revision: https://phab.enlightenment.org/D12090