platform/upstream/efl.git
4 years agodocs: Change non public Tizen Elementary APIs to internal 54/245154/1 accepted/tizen/unified/20201005.101332 submit/tizen/20201005.012108
Myoungwoon Roy, Kim [Mon, 5 Oct 2020 00:03:43 +0000 (09:03 +0900)]
docs: Change non public Tizen Elementary APIs to internal

Change-Id: I19b3afe821c29031a08f473df6f846b7c6a063af

4 years agobuild: addition to abf0e9dffe35da8fa6e20055f53786b2adbec271 accepted/tizen/unified/20200928.072827 submit/tizen/20200927.213403
Marcel Hollerbach [Fri, 25 Sep 2020 12:05:16 +0000 (14:05 +0200)]
build: addition to abf0e9dffe35da8fa6e20055f53786b2adbec271

the fix is also needed in ecore_imf modules.

Thank you Ross!

4 years agoEthumb: reorder _init/shutdown functions, ecore and evas init/shutdown are useless...
Vincent Torri [Fri, 25 Sep 2020 09:30:04 +0000 (10:30 +0100)]
Ethumb: reorder _init/shutdown functions, ecore and evas init/shutdown are useless, as ecore_evas already manages them

Test Plan: compilation

Reviewers: raster

Reviewed By: raster

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12161

4 years agoedje_external_inspector: re-order _init a bit: log after eina, remove ecore_init...
Vincent Torri [Fri, 25 Sep 2020 09:29:54 +0000 (10:29 +0100)]
edje_external_inspector: re-order _init a bit: log after eina, remove ecore_init/shutdown as it is already managed by edje_init/shutdown

Reviewers: raster

Reviewed By: raster

Subscribers: raster, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12160

4 years agobuild: fix dependencies of software engine
Marcel Hollerbach [Thu, 24 Sep 2020 08:28:27 +0000 (10:28 +0200)]
build: fix dependencies of software engine

before the efl-one refactoring ecore_x_deps contained all the
dependencies for x11. After that not anymore, however, we should link
software_x11 to x11 libs directly, which is again happening after this
here.

4 years agoNEWS: fix some typos for 1.25 update
Stefan Schmidt [Wed, 23 Sep 2020 08:15:53 +0000 (10:15 +0200)]
NEWS: fix some typos for 1.25 update

Vicent spotted some typos. Get them fixed here.

4 years agoevas vg: check vg changed 70/244770/1
Shinwoo Kim [Thu, 24 Sep 2020 03:16:49 +0000 (12:16 +0900)]
evas vg: check vg changed

Summary:
It is not able to render even though vg object has a chance to render,
because evas_object_smart_changed_get checks only Evas_Object_Protected_Data,
when Efl_Canvas_Vg_Object_Data.changed is TRUE.

Reviewers: Hermet, jsuya, herb

Reviewed By: Hermet

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12152

Change-Id: Ic318e176e18a97bbb015f2e50008b70ffc565211

4 years agodocs: Changes non public Tizen Elementary APIs to internal 41/244741/1 accepted/tizen/unified/20200928.004556 submit/tizen/20200925.001651
Myoungwoon Roy, Kim [Thu, 24 Sep 2020 01:33:26 +0000 (10:33 +0900)]
docs: Changes non public Tizen Elementary APIs to internal

Change-Id: I29aac5af360035f887406491de850efd50e6c175

4 years agowayland_shm: Check tbm_client validation. 89/244689/1 accepted/tizen/unified/20200924.001846 submit/tizen/20200923.083532
Woochanlee [Wed, 23 Sep 2020 08:05:22 +0000 (17:05 +0900)]
wayland_shm: Check tbm_client validation.

In certain use case, the tbm_client has been deinited when _evas_outbuf_redraws_clear called.
It cause crash.

Change-Id: Ibdc75b3cf66a541e5ac5568e604c54a137f2eb8d

4 years agorelease: Update NEWS and bump version for 1.25.0 release
Stefan Schmidt [Tue, 22 Sep 2020 17:54:20 +0000 (19:54 +0200)]
release: Update NEWS and bump version for 1.25.0 release

4 years agoecore_wl2_window: update window's saved size in xdg_surface_configure callback 00/244600/2
Doyoun Kang [Tue, 22 Sep 2020 08:01:32 +0000 (17:01 +0900)]
ecore_wl2_window: update window's saved size in xdg_surface_configure callback

There was a bug that the window was resized to old size when the window's state changed
to unmaximize and unfullscreen.
This patch updates window's saved size when the window which state is both unmaximize
and unfullscreen.

@tizen_fix

Change-Id: I63decf2113f26f2cfb8378c2a29ff39455ffcf36
Signed-off-by: Doyoun Kang <doyoun.kang@samsung.com>
4 years agoefl_gfx_path: When path end(Z,z), Current point are returned to starting point(M,m) 81/244581/1
JunsuChoi [Tue, 22 Sep 2020 05:52:42 +0000 (14:52 +0900)]
efl_gfx_path: When path end(Z,z), Current point are returned to starting point(M,m)

Summary:
When path ends with 'z' or 'Z' command, if 'm' comes as next command,
the current point is incorrectly referenced.
Since 'Z', 'z' means to close the path,
so, Save point and reuse it when 'M' or 'm' is called.

Test Plan:
[Error Case]
<svg id="Layer_1" data-name="Layer 1" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 64 64">
    <path d="M 52.17,20
             H 11.92
             V 43
             H 52.17
             Z
             m -1.5,21.5
             H 13.42
             v -20
             H 50.67
             Z
" fill="#a8b7c1" stroke="#F00" stroke-width="0.5" />
</svg>

Reviewers: Hermet, smohanty

Reviewed By: Hermet

Subscribers: cedric, herb, kimcinoo, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12158

Change-Id: Ib9104707d3c3297e5cc6a856b3c50c18e288a2d0

4 years agoelm_config: don't prepend /tmp/ to file name
Vincent Torri [Mon, 21 Sep 2020 06:22:40 +0000 (06:22 +0000)]
elm_config: don't prepend /tmp/ to file name

eina_file_mkstemp() already takes care of this.

Differential Revision: https://phab.enlightenment.org/D12155

4 years agoRevert "elm - collection view - check return and fix coverity wanring"
Stefan Schmidt [Mon, 21 Sep 2020 09:06:39 +0000 (11:06 +0200)]
Revert "elm - collection view - check return and fix coverity wanring"

This reverts commit f3fdcf56923c8a2ca126e14d915e6386c7c4ff7e.

This commit leads to ninja test timeout. When returning result here the
future will never fire and we will keep spinning.

4 years agoelm - toiolbar - use correct theme icon for more item
Carsten Haitzler (Rasterman) [Mon, 21 Sep 2020 08:35:06 +0000 (09:35 +0100)]
elm - toiolbar - use correct theme icon for more item

there is a specific thme icon for "more" items when a toolbar
compresses. it literally was not being used. it should use it.
"go-down" that it was using isn't exactly very expressive of what this
item does...

@fix

4 years agoecore - efl thread - dont close invalid < 0 fd's accepted/tizen/unified/20200923.024107 submit/tizen/20200920.213436
Carsten Haitzler (Rasterman) [Sat, 19 Sep 2020 23:06:30 +0000 (00:06 +0100)]
ecore - efl thread - dont close invalid < 0 fd's

fix CID 1396951

4 years agoecore - efl exe - check returns of open and dup2 for stdin etc. fds
Carsten Haitzler (Rasterman) [Sat, 19 Sep 2020 22:59:59 +0000 (23:59 +0100)]
ecore - efl exe - check returns of open and dup2 for stdin etc. fds

check returns - fix CID 1397002

4 years agoeina bench - have default values inside benc dtoa
Carsten Haitzler (Rasterman) [Sat, 19 Sep 2020 22:48:45 +0000 (23:48 +0100)]
eina bench - have default values inside benc dtoa

addess CID 1400856

4 years agoevas test - check return of ftell and malloc and handle properly
Carsten Haitzler (Rasterman) [Sat, 19 Sep 2020 22:43:58 +0000 (23:43 +0100)]
evas test - check return of ftell and malloc and handle properly

fix CID 1400871

4 years agoeina thread - check return of pthread_attr_init
Carsten Haitzler (Rasterman) [Sat, 19 Sep 2020 22:37:22 +0000 (23:37 +0100)]
eina thread - check return of pthread_attr_init

fix CID 1401031

4 years agoemotion - check return of eet data write
Carsten Haitzler (Rasterman) [Sat, 19 Sep 2020 22:34:39 +0000 (23:34 +0100)]
emotion - check return of eet data write

fix CID 1401040

4 years agoethumb - fix idx range check
Carsten Haitzler (Rasterman) [Sat, 19 Sep 2020 22:15:44 +0000 (23:15 +0100)]
ethumb - fix idx range check

fix CID 1402612

4 years agoethumb - fix index range check
Carsten Haitzler (Rasterman) [Sat, 19 Sep 2020 22:13:08 +0000 (23:13 +0100)]
ethumb - fix index range check

fix CID 1402628

4 years agoethumb - fix index range check
Carsten Haitzler (Rasterman) [Sat, 19 Sep 2020 22:11:53 +0000 (23:11 +0100)]
ethumb - fix index range check

fix CID 1402631

4 years agoethumb - check size read
Carsten Haitzler (Rasterman) [Sat, 19 Sep 2020 22:09:05 +0000 (23:09 +0100)]
ethumb - check size read

fix CID 1402633

4 years agoethumb - check inoput index from stdin
Carsten Haitzler (Rasterman) [Sat, 19 Sep 2020 22:03:20 +0000 (23:03 +0100)]
ethumb - check inoput index from stdin

fix CID 1402646

4 years agoevas - psd loader - check read values to try make coverity happy
Carsten Haitzler (Rasterman) [Sat, 19 Sep 2020 21:32:57 +0000 (22:32 +0100)]
evas - psd loader - check read values to try make coverity happy

this is trying to fix CID 1402688, 1402687, 1402682, 1402665. not sure
it will or will not but it does dot i's and cross t's

4 years agoelm - collection view - check return and fix coverity wanring
Carsten Haitzler (Rasterman) [Sat, 19 Sep 2020 19:38:12 +0000 (20:38 +0100)]
elm - collection view - check return and fix coverity wanring

fix CID 1405808

4 years agoecore con url - curl is actually optional and only needs to init on use
Carsten Haitzler (Rasterman) [Mon, 14 Sep 2020 10:09:08 +0000 (11:09 +0100)]
ecore con url - curl is actually optional and only needs to init on use

ecore_con_utl_init is kind of pointless when we can init at the first
use of a url and just have it fail. the problem is anyone initting
ecore_con's url stuff will then pull in curl at that point and not
laer "on first use" which is kind of nw how the dlopen fun was
intended, so push it off until then.

4 years agoelm_config: remove useless init|shutdown of ecore_win32 and eocore_cocoa (done in...
Vincent Torri [Thu, 17 Sep 2020 16:26:59 +0000 (16:26 +0000)]
elm_config: remove useless init|shutdown of ecore_win32 and eocore_cocoa (done in ecore_evas module), and useless wayland variable

ecore_win32 and ecore_cocoa was init. wayland variable was also unused

Reviewed-by: Christopher Michael <devilhorns@comcast.net>
Differential Revision: https://phab.enlightenment.org/D12149

4 years agocall sd_seat_can_tty to determine if ttys should be used, instead of hard-coding...
nerdopolis [Fri, 18 Sep 2020 09:44:23 +0000 (11:44 +0200)]
call sd_seat_can_tty to determine if ttys should be used, instead of hard-coding seat0

Summary: All TTYs are on seat0, but not all seat0s have TTYs

Reviewers: #efl, devilhorns

Reviewed By: #efl, devilhorns

Subscribers: bu5hm4n, stefan_schmidt, raster, cedric, #reviewers, #committers

Tags: #efl

Reviewed-by: Christopher Michael <devilhorns@comcast.net>
Differential Revision: https://phab.enlightenment.org/D12143

4 years agoelm_slider: add A11Y value interface 78/244178/3
Kamil Konieczny [Tue, 15 Sep 2020 11:45:19 +0000 (13:45 +0200)]
elm_slider: add A11Y value interface

Add A11Y value inteface to slider legacy object.

Change-Id: I457270d0abe185e3c57873a92c2052ceeb0675cf

4 years agoefl_ui_widget, elc 46/244346/3 accepted/tizen/unified/20200923.024030 submit/tizen/20200918.044325
Taehyub Kim [Thu, 17 Sep 2020 11:09:09 +0000 (20:09 +0900)]
efl_ui_widget, elc
_popup : disable auto scroll function for popup scroller

Change-Id: I8873c13f4929787ee4c359c7c82a87f45c2dcd64

4 years agoedje_entry: prevent cursor position change in selection state 06/244306/3
Bowon Ryu [Thu, 17 Sep 2020 06:05:29 +0000 (15:05 +0900)]
edje_entry: prevent cursor position change in selection state

this patch prevents the cursor position from change due to mouse up on selection state

Change-Id: I57458f72aad1e5be0afdde1be7664c2e48004c26
Signed-off-by: Bowon Ryu <bowon.ryu@samsung.com>
4 years agoevas_outbuf: Fix memory leak.
Woochanlee [Thu, 17 Sep 2020 08:23:41 +0000 (17:23 +0900)]
evas_outbuf: Fix memory leak.

Summary: In certain use cases, when image data is created twice, the existing memory pointer is isolated and a memory leak occurs.

Reviewers: raster, Hermet

Reviewed By: Hermet

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8813

Differential Revision: https://phab.enlightenment.org/D12148

4 years agoevas_main: Fix memory leak.
Woochanlee [Thu, 17 Sep 2020 08:23:10 +0000 (17:23 +0900)]
evas_main: Fix memory leak.

Summary: If the inlist has only one item when it removed. the data is not freed in descturctor.

Reviewers: Hermet, raster, vtorri

Reviewed By: Hermet

Subscribers: vtorri, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12147

4 years agodocs: Corrects EFL main pages and Changes non public Tizen Elementary APIs to internal 11/244311/2
Myoungwoon Roy, Kim [Thu, 17 Sep 2020 06:39:35 +0000 (15:39 +0900)]
docs: Corrects EFL main pages and Changes non public Tizen Elementary APIs to internal

Change-Id: I8442e25725aa28222bef6f6de75d57dc28ccc2c0

4 years agoevas vg: check object changed as well 2 88/244288/1
Shinwoo Kim [Thu, 17 Sep 2020 01:40:24 +0000 (10:40 +0900)]
evas vg: check object changed as well 2

Summary:
evas cannot render vg object if Efl_Canvas_Vg_Node_Data flag is not
EFL_GFX_CHANGE_FLAG_NONE and Evas_Object_Protected_Data.changed is FALSE,
when vg object marks its node as changed.

Above case could be possible if vg object render_pre is not called, and
only nd->vd->obj->changed is set to false by evas_object_change_reset.

Reviewers: Hermet, jsuya, herb

Reviewed By: Hermet

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12142

Change-Id: I628b3596c282354005ae2a36209460d68bb43c01

4 years agoevas_textblock: prevent textnodes with <ps>/<br> without format node
Ali Alzyod [Mon, 14 Sep 2020 16:30:06 +0000 (16:30 +0000)]
evas_textblock: prevent textnodes with <ps>/<br> without format node

this will prevent textnodes content with <ps> or <br> without format node

Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D12145

4 years agoedje_entry: prevent incorrect cursor position during preedit 35/244035/3 accepted/tizen/unified/20200917.030601 submit/tizen/20200916.073812
Bowon Ryu [Mon, 14 Sep 2020 07:51:03 +0000 (16:51 +0900)]
edje_entry: prevent incorrect cursor position during preedit

due to the influence of elm_entry_filter_limit_size,
"cursor_pos" from imf and "end_pos - start_pos" may be different.
in particular, if the preedit string can not be entered at all due to filter,
cursor_pos is larger than 0, so the cursor moves to an unintended position.

This patch avoids the issue by moving the cursor's position to end_pos in this case.

@tizen_fix

Change-Id: Ie7ad48ad02ecbf26a0bedab960cd0bf9042a5995
Signed-off-by: Bowon Ryu <bowon.ryu@samsung.com>
4 years agovg_load_svg: Support dasharray attribute for stroke 44/244144/1
JunsuChoi [Tue, 15 Sep 2020 03:11:49 +0000 (12:11 +0900)]
vg_load_svg: Support dasharray attribute for stroke

Summary:
It supports stroke-dasharray, one of the stroke properties of svg.
https://developer.mozilla.org/en-US/docs/Web/SVG/Attribute/stroke-dasharray

Test Plan:
[Test SVG]
<svg viewBox="0 0 30 10" xmlns="http://www.w3.org/2000/svg">
  <!-- No dashes nor gaps -->
  <line x1="0" y1="1" x2="30" y2="1" stroke="black" />

  <!-- Dashes and gaps of the same size -->
  <line x1="0" y1="3" x2="30" y2="3" stroke="black"
          stroke-dasharray="4" />

  <!-- Dashes and gaps of different sizes -->
  <line x1="0" y1="5" x2="30" y2="5" stroke="black"
          stroke-dasharray="4 1" />

  <!-- Dashes and gaps of various sizes with an odd number of values -->
  <line x1="0" y1="7" x2="30" y2="7" stroke="black"
          stroke-dasharray="4 1 2" />

  <!-- Dashes and gaps of various sizes with an even number of values -->
  <line x1="0" y1="9" x2="30" y2="9" stroke="black"
          stroke-dasharray="4 1 2 3" />
</svg>

Reviewers: Hermet, smohanty

Reviewed By: Hermet

Subscribers: kimcinoo, herb, cedric, #committers, #reviewers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12146

Change-Id: I05d6a302b7f9546e68bd393da1890869cf599e6e

4 years agorelease: Update NEWS and bump version for 1.25.0-beta3 release
Stefan Schmidt [Mon, 14 Sep 2020 10:12:31 +0000 (12:12 +0200)]
release: Update NEWS and bump version for 1.25.0-beta3 release

4 years agoelm_dnd: clean up the registered target object when elm_drop_target_del API is called 55/244055/2 accepted/tizen/unified/20200915.014603 submit/tizen/20200914.112517
Taehyub Kim [Mon, 14 Sep 2020 10:48:27 +0000 (19:48 +0900)]
elm_dnd: clean up the registered target object when elm_drop_target_del API is called

@fix

Change-Id: Idbe9d1412093c25e3fa4b20e537d17f70ac5556b

4 years agoEfl.Canvas.Vg: When VG container has transparency, ector buffer created synchronously 41/243941/2
JunsuChoi [Fri, 11 Sep 2020 08:02:23 +0000 (17:02 +0900)]
Efl.Canvas.Vg: When VG container has transparency, ector buffer created synchronously

Summary:
If the container has transparency, it internally alpha blends with ector buffer.
So ector buffer must be created synchronously.

Test Plan: N/A

Reviewers: Hermet, smohanty

Reviewed By: Hermet

Subscribers: #reviewers, #committers, cedric, herb, kimcinoo

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12140

Change-Id: I00c8560e9fb8f021cba219dd8105dda90dc8deed

4 years agoui_widget: restore hide callback for sub-object 37/243937/1 accepted/tizen/unified/20200914.131356 submit/tizen/20200913.215058
Yeongjong Lee [Fri, 11 Sep 2020 08:21:19 +0000 (17:21 +0900)]
ui_widget: restore hide callback for sub-object

In commit 8d47ced0497c7d24203a16c5011c57928822024e, hide callback to restore
focus was removed. this patch restores the hide callback to fix focus issue
when hiding a focused widgets.

@tizen_fix

Change-Id: I07cd3bcc329316a03a6e418141c11403a7d49d0d

4 years agovg_load_svg: Add Mask, ClipPath nodees as dummy. 27/243927/1
JunsuChoi [Fri, 11 Sep 2020 02:49:09 +0000 (11:49 +0900)]
vg_load_svg: Add Mask, ClipPath nodees as dummy.

Summary:
Mask and clippath TAG are not supported yet.
If this TAG is used, the Child TAG declared with this TAG may cause problems.
To prevent that, declare them as Unknown types until they are supported. (display=none)

Test Plan: N/A

Reviewers: Hermet, smohanty

Reviewed By: Hermet

Subscribers: cedric, herb, #reviewers, kimcinoo, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12139

Change-Id: I007c4538d9b2ab92b6b58e954215c63a23541d18

4 years agovg_load_svg: Fix missing loader.gradients update 26/243926/1
JunsuChoi [Fri, 11 Sep 2020 02:48:30 +0000 (11:48 +0900)]
vg_load_svg: Fix missing loader.gradients update

Summary:
When both <defs> and <svg> gradients are declared,
the update of gradient declared in is omitted.
Therefore, this patch fixes any missing gradient updates.

Test Plan: N/A

Reviewers: Hermet, smohanty

Reviewed By: Hermet

Subscribers: #reviewers, #committers, cedric, herb, kimcinoo

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12138

Change-Id: I82cb746931484041cc0662a5ad770dce1efcc1d2

4 years agoEfl.Canvas.Vg.Object: Fix typo 25/243925/1
JunsuChoi [Fri, 11 Sep 2020 02:46:58 +0000 (11:46 +0900)]
Efl.Canvas.Vg.Object: Fix typo

Summary:
The 3th parameter of ector_buffer_pixels_get is the width.
So chnage px to pw.

Test Plan: N/A

Reviewers: Hermet, smohanty

Reviewed By: Hermet

Subscribers: cedric, herb, #reviewers, kimcinoo, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12141

Change-Id: I05555f5583bb5a8cadc7bff25422af01731befa8

4 years agoelm_spinner: Fix button and entry show/hide transition timing. 09/243909/1 submit/tizen/20200911.065559
Woochanlee [Fri, 11 Sep 2020 06:39:05 +0000 (15:39 +0900)]
elm_spinner: Fix button and entry show/hide transition timing.

If the signal emiited after show/hide API calling,
internal edje calc logic will broken.

@tizen_fix

Change-Id: Ib5fa1a4bfeea0867f27faeddd00f3e0e9ccdbbd3
Signed-off-by: Woochanlee <wc0917.lee@samsung.com>
4 years agodocs: Correct the mismatched group name in Evas.
Myoungwoon Roy, Kim [Mon, 7 Sep 2020 08:49:53 +0000 (08:49 +0000)]
docs: Correct the mismatched group name in Evas.

I had found mismatched group name in Evas and fixed them.

Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D12134

4 years agovg_load_svg: Supports case when only rx or ry is declared
JunsuChoi [Thu, 10 Sep 2020 05:59:29 +0000 (14:59 +0900)]
vg_load_svg: Supports case when only rx or ry is declared

Summary:
In relation to the declaration of rx and ry attribute of rect, the following three cases occur.
rx="10" (or ry="10"
rx="10" ry = "0" (or rx="0" ry = "10")
rx="10" ry = "10"
To cover these case, we check the rx and ry declarations.

Test Plan: N/A

Reviewers: Hermet, smohanty

Subscribers: kimcinoo, herb, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12137

4 years agoevas vg: check object changed as well
Shinwoo Kim [Wed, 9 Sep 2020 11:14:48 +0000 (12:14 +0100)]
evas vg: check object changed as well

Summary:
evas cannot render vg object if Efl_Canvas_Vg_Object_Data.changed is ture
and Evas_Object_Protected_Data.changed is false, when vg object marks itself
as changed.

Above case is possible depending on the draw area which is calculated by
eng_output_redraws_next_update_get. If this function returns NULL,
the vg object render function is not called, and vd->changed remains true.
Only vd->obj->changed is set to false by vas_object_change_reset.

Reviewers: Hermet, jsuya, herb, raster

Reviewed By: raster

Subscribers: raster, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12136

4 years agoecore_evas_wayland_common: check device class when device is added. 98/243598/2
Hosang Kim [Tue, 8 Sep 2020 10:32:19 +0000 (19:32 +0900)]
ecore_evas_wayland_common: check device class when device is added.

Change-Id: I1d73f9cbf34bb46a01f2428dda4cf9443187c508

4 years agoevas/canvas/vg: Added thorvg initialization. 14/241114/4
Michal Szczecinski [Fri, 7 Aug 2020 07:04:42 +0000 (09:04 +0200)]
evas/canvas/vg: Added thorvg initialization.

Added thorvg initialization in evas_vg_cache. To don't break current
code, added separated functions to render object when thorvg lib is
used.

Change-Id: I2ea361fab9758a170f971dc7a731eded9d45b0e6

4 years agoelm_entry: legacy smart selection[start,cleared] callback fix 62/243562/1
Ali Alzyod [Tue, 8 Sep 2020 06:55:30 +0000 (15:55 +0900)]
elm_entry: legacy smart selection[start,cleared] callback fix

Summary:
due to changes in latest text apis in unified, legacy selection call back [start,clear] does not work anymore,
Now add the support for legacy callback [selection,start   selection,cleared]

Reviewers: woohyun, bowonryu, stefan_schmidt

Reviewed By: bowonryu

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12126

Change-Id: Ib23e07470d3ff28caecc700d26047c3245d28f8a

4 years agoedje: resolve the possible use of uninitialized value 91/243391/2
Jongmin Lee [Mon, 7 Sep 2020 08:31:26 +0000 (17:31 +0900)]
edje: resolve the possible use of uninitialized value

Change-Id: I5243bb5a0a703812f7b4ef02642f6f0f669e7933

4 years agoCreated gradient manipulation example 99/243499/2
Michal Maciola [Mon, 24 Aug 2020 09:30:52 +0000 (11:30 +0200)]
Created gradient manipulation example

Change-Id: I2f58fba0f6260cf57dbcc58aa59ba26a80eb24ce

4 years agorelease: Update NEWS and bump version for 1.25.0-beta2 release
Stefan Schmidt [Mon, 7 Sep 2020 10:54:04 +0000 (12:54 +0200)]
release: Update NEWS and bump version for 1.25.0-beta2 release

4 years agoelm_gesture_layer: fix crash caused by elm_gesture_layer_cb_del in cb
Jaehyun Cho [Mon, 7 Sep 2020 05:35:12 +0000 (14:35 +0900)]
elm_gesture_layer: fix crash caused by elm_gesture_layer_cb_del in cb

elm_gesture_layer_cb_del can be called in Elm_Gesture_Event_Cb cb set by
elm_gesture_layer_cb_set and cb_info of the cb is free in the cb.

This causes crash if cb_info is searched with EINA_INLIST_FOREACH.

4 years agoRevert "atspi: get_at_point and neighbor navigation unified" 19/243319/3
Kamil Konieczny [Fri, 4 Sep 2020 12:31:19 +0000 (14:31 +0200)]
Revert "atspi: get_at_point and neighbor navigation unified"

When there is floating button on screen, it is not available
for atspi highlight on OneFingerSingleTap. Reverting this commit
fixes it.

This reverts commit 98d14ca80d3d5540760ff41d85f2c61e7bd40c6b

Change-Id: If4204374b081e8711e17e37f9f23081c7486e815

4 years agoelm cnp - make matchers const and static too accepted/tizen/unified/20200907.045103 submit/tizen/20200906.212734
Carsten Haitzler (Rasterman) [Sat, 5 Sep 2020 14:03:44 +0000 (15:03 +0100)]
elm cnp - make matchers const and static too

4 years agoelm_cnp: adjust image type to real format
Marcel Hollerbach [Sat, 5 Sep 2020 07:07:12 +0000 (09:07 +0200)]
elm_cnp: adjust image type to real format

we should never say that we have image/png etc. if we do not have the
respective format. Otherwise we are just confusing the other client.

With this you can copy things from elm to thunderbird / libreoffice
writer / firefox / telegram / chromium.

Please note if you test this: Start *a new xorg session* it was
discovered that xorg clients tent to be totally confused if there was a
protocol error before, or some wrong mime type was delivered before,
thunderbird even likes to crash if this is not done.

This should at some point be extended to more image types.

4 years agoecore_evas_x: reply in *any* way with XdndStatus
Marcel Hollerbach [Fri, 4 Sep 2020 16:22:49 +0000 (18:22 +0200)]
ecore_evas_x: reply in *any* way with XdndStatus

otherwise Xorg stops sometimes to continue sending callbacks to us.
Which results in a broken state maschine.

4 years agoecore x - fix typo in internal backdoor api to get vsync debug
Carsten Haitzler (Rasterman) [Fri, 4 Sep 2020 13:39:15 +0000 (14:39 +0100)]
ecore x - fix typo in internal backdoor api to get vsync debug

4 years agoecore drm - increase timeout for flips from 0.05 to 2.0 sec
Carsten Haitzler (Rasterman) [Fri, 4 Sep 2020 12:04:58 +0000 (13:04 +0100)]
ecore drm - increase timeout for flips from 0.05 to 2.0 sec

so our timeout is far too agressive. we cant stay hung on buggy
devices/drivers though so make the timeout 2s instead of 0.05s so we
dont go retrying flips too agressively which causes stuttering

@fix

4 years agovg_load_svg: Fix pointing to wrong parent
JunsuChoi [Fri, 4 Sep 2020 02:14:58 +0000 (11:14 +0900)]
vg_load_svg: Fix pointing to wrong parent

Summary:
If defs type is open empty style, it is not included in stack.
Because the current logic is weak to group tags in sibling relationships.
If defs of open-empty type is placed before the group tag,
the group tag refers to the wrong parent. This patch prevents it.

This change is the same concept as the D12127, which was reverted to 5065c9d0c2.

Test Plan: N/A

Reviewers: Hermet, smohanty

Reviewed By: Hermet

Subscribers: cedric, herb, kimcinoo, #committers, #reviewers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D12129

4 years agoevas/textblock: fix a memory issue caused by recreating ellipsis item 52/243252/2
Youngbok Shin [Fri, 4 Sep 2020 08:18:06 +0000 (17:18 +0900)]
evas/textblock: fix a memory issue caused by recreating ellipsis item

Change-Id: Ib50e95a5f72156afc4d4112cb3a9b3f0de2fdae0

4 years agodocs: Add the missing API page name to elementary and evas
Myoungwoon Roy, Kim [Thu, 3 Sep 2020 05:33:52 +0000 (05:33 +0000)]
docs: Add the missing API page name to elementary and evas

I had found missing API reference page name. There should be page name to each API defgroup

Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D12128

4 years agoedje - calc - handle swallow disappearing mid-calc due to callabcks
Carsten Haitzler (Rasterman) [Thu, 3 Sep 2020 08:41:35 +0000 (09:41 +0100)]
edje - calc - handle swallow disappearing mid-calc due to callabcks

callbacks attached to move/resize etc. can change a swallow and ..
well.. i hit a crash where it happens. so handle this case.

@fix

4 years agoRevert "SvgLoader: Defs type nodes are not saved in loader's node list."
Hermet Park [Thu, 3 Sep 2020 07:45:27 +0000 (16:45 +0900)]
Revert "SvgLoader: Defs type nodes are not saved in loader's node list."

This reverts commit 40b6e44947740133df20d97ca3ed65268037af20.

This has some side effects, we will intrdouce it again after fix it.

4 years agoecore_wl2: remove hard-coded tizen-policy-ext protocol 72/242572/2
Junseok, Kim [Fri, 28 Aug 2020 04:30:03 +0000 (13:30 +0900)]
ecore_wl2: remove hard-coded tizen-policy-ext protocol

Use tizen-policy-ext protocol in wayland-extension instead of hard-coded protocol

Change-Id: Ic287f62bea5b7a223d15719538970c73fc45d409

4 years agoelm_config: Fix wrong type conversion 85/243085/1
Woochanlee [Thu, 3 Sep 2020 05:01:25 +0000 (14:01 +0900)]
elm_config: Fix wrong type conversion

User set the env value as "0" or "1"
but in previouse case the result was always true.

Change-Id: If51ce58070e125141f51e10515330dc7bc61e69e

4 years agoSvgLoader: Defs type nodes are not saved in loader's node list.
JunsuChoi [Wed, 2 Sep 2020 09:34:59 +0000 (18:34 +0900)]
SvgLoader: Defs type nodes are not saved in loader's node list.

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

4 years agoeina file test: check rmdir result.
Vincent Torri [Wed, 2 Sep 2020 08:21:39 +0000 (08:21 +0000)]
eina file test: check rmdir result.

Fix CID 1432267

Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D12125

4 years agobuild: only find env once
Marcel Hollerbach [Wed, 2 Sep 2020 07:25:51 +0000 (09:25 +0200)]
build: only find env once

this was a bit messy, and we might have caused a bug on the ML with
that.

4 years agoecore_wl2: check reference count for avoiding crash 66/242966/1
Hosang Kim [Wed, 2 Sep 2020 04:16:09 +0000 (13:16 +0900)]
ecore_wl2: check reference count for avoiding crash

@tizen_only
Change-Id: Icbbcfa21a309afef7d871981d94f82dc4ab5b9b0

4 years agoevas_vg: remove since tizen 6.0 tags for auto generating tags 91/242291/2
Taehyub Kim [Tue, 25 Aug 2020 10:27:40 +0000 (19:27 +0900)]
evas_vg: remove since tizen 6.0 tags for auto generating tags

Change-Id: I692b25e3044a509581979a2400d2ce5784f1fac4

4 years agoelm_config: Fix wrong type conversion 20/242920/1
Woochanlee [Wed, 2 Sep 2020 03:04:31 +0000 (12:04 +0900)]
elm_config: Fix wrong type conversion

User set the env value as "0" or "1"
but in previouse case the result was always true.

Change-Id: I97206cbc06d73211c871b4d5cb9e44d54ef85b31

4 years agoefl_ui_focus_manager_calc: fix CID submit/tizen/20200901.214039
Marcel Hollerbach [Tue, 1 Sep 2020 12:10:21 +0000 (14:10 +0200)]
efl_ui_focus_manager_calc: fix CID

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.

4 years agoeldbus: use the const macros we have
Marcel Hollerbach [Tue, 1 Sep 2020 11:25:06 +0000 (13:25 +0200)]
eldbus: use the const macros we have

clang warns about that, its rather harmless, but nicer.

4 years agoefl: resolve warnings over misleading code
Marcel Hollerbach [Tue, 1 Sep 2020 11:23:42 +0000 (13:23 +0200)]
efl: resolve warnings over misleading code

this is meant to be like this, but clang thought of an intention mistake
here that went unsighted.

4 years agobenchmarks: specify this outside extern "C"
Marcel Hollerbach [Tue, 1 Sep 2020 11:10:03 +0000 (13:10 +0200)]
benchmarks: specify this outside extern "C"

uint128 is defined as a c++ specific type.

4 years agoeldbus_fake_server: fix string format usage
Marcel Hollerbach [Tue, 1 Sep 2020 11:03:46 +0000 (13:03 +0200)]
eldbus_fake_server: fix string format usage

we know that there is no % etc. formatting in this string, however, we
can just use the ck abort msg macro from libcheck directly.

4 years agoefl: remove unused functions
Marcel Hollerbach [Tue, 1 Sep 2020 11:09:37 +0000 (13:09 +0200)]
efl: remove unused functions

these are not used, if they are used again in the future, they can be
get back via git.

4 years agoefl: remove bool and string pointer mixup
Marcel Hollerbach [Tue, 1 Sep 2020 11:02:51 +0000 (13:02 +0200)]
efl: remove bool and string pointer mixup

4 years agoelm icon/thumb - fix handling of a wider range of video extns
Carsten Haitzler (Rasterman) [Tue, 1 Sep 2020 11:25:01 +0000 (12:25 +0100)]
elm icon/thumb - fix handling of a wider range of video extns

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

4 years agoevas_textblock: remove style padding from native width and formatted height
Youngbok Shin [Tue, 1 Sep 2020 10:39:13 +0000 (19:39 +0900)]
evas_textblock: remove style padding from native width and formatted height

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

4 years agoeeze_mount binary: more secure check extension ('.iso' and not 'iso')
Vincent Torri [Mon, 31 Aug 2020 13:54:53 +0000 (13:54 +0000)]
eeze_mount binary: more secure check extension ('.iso' and not 'iso')

Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D12123

4 years agoecore_file_can_exec: fix .bat case
Vincent Torri [Mon, 31 Aug 2020 13:46:51 +0000 (13:46 +0000)]
ecore_file_can_exec: fix .bat case

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

4 years agoelm_entry: Remove drag & drop feature in entry. 11/242811/2
Woochanlee [Tue, 1 Sep 2020 04:21:24 +0000 (13:21 +0900)]
elm_entry: Remove drag & drop feature in entry.

The drag & drop feature is no more support in Tizen platform.
there is no UX for it.

@tizen_feature

Change-Id: Ia0956943cc8ccc94add9cfea30063b774cc79795

4 years agorelease: Update NEWS and bump version for 1.25.0-beta1 release
Stefan Schmidt [Mon, 31 Aug 2020 12:53:15 +0000 (14:53 +0200)]
release: Update NEWS and bump version for 1.25.0-beta1 release

4 years agoEvil: add gettimeofday()
Vincent Torri [Wed, 26 Aug 2020 18:44:28 +0000 (18:44 +0000)]
Evil: add gettimeofday()

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

4 years agoEcore: more precise implementation of ecore_time_get on Windows
Vincent Torri [Sat, 22 Aug 2020 05:56:05 +0000 (05:56 +0000)]
Ecore: more precise implementation of ecore_time_get on Windows

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

4 years agoecore-file: fix ecore_file_can_exec() on Windows
Vincent Torri [Mon, 31 Aug 2020 11:55:09 +0000 (11:55 +0000)]
ecore-file: fix ecore_file_can_exec() on Windows

on Windows access() has no support of X_OK (see https://docs.microsoft.com/en-us/cpp/c-runtime-library/reference/access-waccess?view=vs-2019), use SHGetFileInfo() instead

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/D12118

4 years agoeina: Replace remove with rmdir and delete tmpstr
João Paulo Taylor Ienczak Zanette [Mon, 24 Aug 2020 16:52:43 +0000 (16:52 +0000)]
eina: Replace remove with rmdir and delete tmpstr

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

4 years agoelm_theme: more secure check extension as '.edj' and not 'edj'
Vincent Torri [Sun, 30 Aug 2020 20:22:49 +0000 (20:22 +0000)]
elm_theme: more secure check extension as '.edj' and not 'edj'

Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Differential Revision: https://phab.enlightenment.org/D12121

4 years agoevas load - dont lose skip flg update after async head skip load fix submit/tizen/20200830.214225
Carsten Haitzler (Rasterman) [Sun, 30 Aug 2020 17:46:48 +0000 (18:46 +0100)]
evas load - dont lose skip flg update after async head skip load fix

119d9f39dd9ba67068a6abb77990ea859d4266f6 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

4 years agoevas - image cache - fix loading with skip head on to load changed files
Carsten Haitzler (Rasterman) [Sun, 30 Aug 2020 11:05:17 +0000 (12:05 +0100)]
evas - image cache - fix loading with skip head on to load changed files

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

4 years agoeina - statgen (stat generation) - fix enable api to actually enable
Carsten Haitzler (Rasterman) [Sun, 30 Aug 2020 11:04:13 +0000 (12:04 +0100)]
eina - statgen (stat generation) - fix enable api to actually enable

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

4 years agoecore exe - fig signal block to unblock again - some sys dont on exec
Carsten Haitzler (Rasterman) [Fri, 28 Aug 2020 12:48:20 +0000 (13:48 +0100)]
ecore exe - fig signal block to unblock again - some sys dont on exec

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