platform/upstream/efl.git
4 years agoevas_gl_font: free data paths if alloca fails
Ali Alzyod [Tue, 10 Mar 2020 02:41:57 +0000 (11:41 +0900)]
evas_gl_font: free data paths if alloca fails

Reviewers: Hermet, woohyun, bu5hm4n, zmike

Reviewed By: Hermet

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11473

4 years agoecore_evas: do not leak allocated memory in error case
Marcel Hollerbach [Mon, 9 Mar 2020 16:07:35 +0000 (17:07 +0100)]
ecore_evas: do not leak allocated memory in error case

4 years agoecore_evas: only free content when content is != NULL
Marcel Hollerbach [Mon, 9 Mar 2020 16:05:17 +0000 (17:05 +0100)]
ecore_evas: only free content when content is != NULL

otherwise we might run into issues with that.

4 years agoecore_evas_x: fix memory leak when enter drag and drop, without a ee
Marcel Hollerbach [Mon, 9 Mar 2020 16:00:29 +0000 (17:00 +0100)]
ecore_evas_x: fix memory leak when enter drag and drop, without a ee

this is something that can *never* happen otherwise we would have never
got this call. But lets make coverity happy.

4 years agoefl_ui_textbox: dont leak a array when no type is available
Marcel Hollerbach [Mon, 9 Mar 2020 15:59:33 +0000 (16:59 +0100)]
efl_ui_textbox: dont leak a array when no type is available

this fixes a possible leak.

4 years agoecore_evas_x: correctly handle images in X11
Marcel Hollerbach [Mon, 9 Mar 2020 15:47:16 +0000 (16:47 +0100)]
ecore_evas_x: correctly handle images in X11

I thought that i explicitly tested this. However, it seems i was wrong,
this way now this is handled correctly, and the image is stored and
coverted to a path.

With this you can now copy images from chromium and firefox to elm apps.

The change in elm_entry reverts parts of the rewrite of the elm
handling, as initially every path came *without* "file://" in front, so
we have to maintain that.

fixes T8625

4 years agoelm_cnp: move free call after the last usage
Marcel Hollerbach [Mon, 9 Mar 2020 15:46:01 +0000 (16:46 +0100)]
elm_cnp: move free call after the last usage

this is however pretty much irrelevant as only the pointer is used.

4 years agoelm_cnp: free array when going out of scope
Marcel Hollerbach [Mon, 9 Mar 2020 15:45:36 +0000 (16:45 +0100)]
elm_cnp: free array when going out of scope

this was missing and indeed a memory leak.

4 years agoefl_ui_win: Fix null pointer dereference
Christopher Michael [Mon, 9 Mar 2020 13:07:48 +0000 (09:07 -0400)]
efl_ui_win: Fix null pointer dereference

Fixes Coverity CID1420321

4 years agoefl_ui_win: Fix null pointer dereference
Christopher Michael [Mon, 9 Mar 2020 13:03:58 +0000 (09:03 -0400)]
efl_ui_win: Fix null pointer dereference

Fixes Coverity CID1420329 and CID1420328

4 years agoefl_ui_win: Fix horrible formatting
Christopher Michael [Mon, 9 Mar 2020 13:02:21 +0000 (09:02 -0400)]
efl_ui_win: Fix horrible formatting

NB: No functional changes

4 years agoeina_content: fix use after free in test code
Marcel Hollerbach [Mon, 9 Mar 2020 12:05:38 +0000 (13:05 +0100)]
eina_content: fix use after free in test code

4 years agoeina_content: free memory in error paths
Marcel Hollerbach [Mon, 9 Mar 2020 12:04:35 +0000 (13:04 +0100)]
eina_content: free memory in error paths

4 years agoeina: fix define of Eina_Content
Marcel Hollerbach [Mon, 9 Mar 2020 09:03:37 +0000 (10:03 +0100)]
eina: fix define of Eina_Content

there is no need to declare that const.

4 years agodoxygen docs: remove @ender_group and organize eina vectors
Xavi Artigas [Mon, 9 Mar 2020 12:01:11 +0000 (13:01 +0100)]
doxygen docs: remove @ender_group and organize eina vectors

Mysterious @ender_group is an invalid doxygen command and I could not find
it defined anywhere. I could not infer its purpose either so... removed.
Put the Eina Vector 2D and 3D groups in their proper place so they appear
with the rest of the Eina basic types.

4 years agodoxygen docs: Remove invalid comments
Xavi Artigas [Mon, 9 Mar 2020 10:55:22 +0000 (11:55 +0100)]
doxygen docs: Remove invalid comments

They contain no information whatsoever and they contain invalid doxygen commands.
This was probably copied over from some other documentation format.

4 years agodoxygen docs: Move smart object methods to proper group
Xavi Artigas [Mon, 9 Mar 2020 10:43:33 +0000 (11:43 +0100)]
doxygen docs: Move smart object methods to proper group

Move smart object methods from @ingroup Evas_Canvas to Evas_Smart_Object_Group
Also, remove lots of redundant @ingroup Evas_Smart_Object_Group

4 years agodoxygen docs: remove doubly-defined @examples
Xavi Artigas [Mon, 9 Mar 2020 09:24:22 +0000 (10:24 +0100)]
doxygen docs: remove doubly-defined @examples

This index page only contains references (@ref), the actual
examples (@example) are defined elsewhere and the double
definition was throwing warnings.

4 years agoRevert "ci: travis: stop osx builds until build failures are solved on Travis"
Stefan Schmidt [Mon, 9 Mar 2020 08:28:12 +0000 (09:28 +0100)]
Revert "ci: travis: stop osx builds until build failures are solved on Travis"

This was actually a problem we could solve on our side. Marcel has done
so in commit: 32597e7e7784049927291e6d8ad131a2b8227d0e

This reverts commit b554b25c4cfd5b49d7ea8b9daf9c52977a68805e.

4 years agoeina: Docs/Comments: eina_safety_checks: Correct grammar, spelling mistakes and be...
Xavi Artigas [Mon, 9 Mar 2020 08:58:19 +0000 (09:58 +0100)]
eina: Docs/Comments: eina_safety_checks: Correct grammar, spelling mistakes and be more clearer

Summary: This just corrects the comments grammar, spelling mistakes and is a little bit clearer.

Reviewers: segfaultxavi, barbieri

Reviewed By: segfaultxavi

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11472

4 years agodoxygen docs: Miscellaneous typos
Xavi Artigas [Thu, 5 Mar 2020 08:35:05 +0000 (09:35 +0100)]
doxygen docs: Miscellaneous typos

4 years agoreadme - adapt for removal of xine/vlc support
Carsten Haitzler (Rasterman) [Sun, 8 Mar 2020 12:38:58 +0000 (12:38 +0000)]
readme - adapt for removal of xine/vlc support

4 years agoecore_evas_x: honor forign dnd implementations
Marcel Hollerbach [Sun, 8 Mar 2020 11:25:35 +0000 (12:25 +0100)]
ecore_evas_x: honor forign dnd implementations

enlightenment does not use the elm dnd operation implementations.
Therefore we have to ensure that we only handle these operations here,
when ecore_evas dnd API is used.

4 years agoecore_evas: Use EFL naming convention in cnp & dnd methods
Xavi Artigas [Tue, 3 Mar 2020 12:10:49 +0000 (13:10 +0100)]
ecore_evas: Use EFL naming convention in cnp & dnd methods

Some methods were missing the "Drag" or "Selection" namespaces or the _Cb suffix.
Depends on D11219

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11426

4 years agomono-tests: Fix build after DnD changes
Xavi Artigas [Tue, 28 Jan 2020 17:33:42 +0000 (17:33 +0000)]
mono-tests: Fix build after DnD changes

The test struct used to check mono bindings struct pack/unpack
is no longer available. Used a different one, and checked the same things.

Reviewed-by: João Paulo Taylor Ienczak Zanette <joao.tiz@expertisesolutions.com.br>
Differential Revision: https://phab.enlightenment.org/D11219

4 years agodocs: Strengthen docs for Copy&Paste and Drag&Drop
Xavi Artigas [Tue, 3 Mar 2020 10:12:08 +0000 (11:12 +0100)]
docs: Strengthen docs for Copy&Paste and Drag&Drop

Including Eina.Content

And a typo/bugfix in ecore_evas_x.

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11204

4 years agoport cnp on Windows
Marcel Hollerbach [Sun, 8 Mar 2020 09:46:09 +0000 (10:46 +0100)]
port cnp on Windows

Test Plan: Ctrl-c and Ctrl-Vworking

Reviewers: raster, zmike

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11439

4 years agoecore_evas: Introduce cnp support for cocoa
Marcel Hollerbach [Fri, 14 Feb 2020 13:10:44 +0000 (14:10 +0100)]
ecore_evas: Introduce cnp support for cocoa

with this commit you can do limited cnp for cocoa. You still cannot copy
and paste pictures or markup arround, only text is supported so far.
However, text on its own works quite stable and good.

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11351

Change-Id: Iec8d57633832d5a213565237f603e3c7f1b1efb2

4 years agoecore_cocoa: change clipboard API
Marcel Hollerbach [Fri, 14 Feb 2020 13:09:21 +0000 (14:09 +0100)]
ecore_cocoa: change clipboard API

the new API works with mimetypes, so we can remove the cnp types from
Ecore_Cocoa.h and just forward the types from ecore_evas directly

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11350

4 years agorewrite efl cnp and dnd handling
Marcel Hollerbach [Sun, 19 Jan 2020 12:58:26 +0000 (13:58 +0100)]
rewrite efl cnp and dnd handling

the previous commits introduced a abstraction for drag in drop which can
be now used for this here. With this commit all the direct protocol
handling in efl.ui is removed, and only the ecore evas API is used.

Additionally, this lead to a giant refactor of how APIs do work. All
Efl.Ui. interfaces have been removed except Efl.Ui.Selection and
Efl.Ui.Dnd, these two have been restructored.
A small list of what is new:
- In general no function pointers are used anymore. They feel very
  uncompftable in bindings and in C. For us its a lot easier to just
listen to a event when a drop enters or leaves, there is no need to
register custom functions for that.
- Asynchronous data transphere is handled via futures, which proved to
  be more error safe.
- Formats and actions are handled as mime types / strings.
- 0 is the default seat if you do not know what else to take.
- Content is in general passes as a content container from eina, this
  also allows applications to pass custom types

The legacy dnd and cnp API is implemented based on that.
All cnp related things are in elm_cnp.c the dnd parts are in elm_dnd.c

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11190

4 years agoecore_evas: introduce wayland support for cnp & dnd
Marcel Hollerbach [Mon, 10 Feb 2020 08:20:18 +0000 (09:20 +0100)]
ecore_evas: introduce wayland support for cnp & dnd

This adds cnp support, actions are right now only mapped to "ask",
further support can be added there, and synchronization can be added to
register more available actions. However, i did not find *any* wayland
implementation in gtk qt nor chromiumos that even use the action to
indicate anything. This here also has a slightly different behaviour to
X11 in terms of coordinates for motion,leave,enter. They can contain
negative coordinates (which is due to the fact that wl is CSD and X11 is
SSD. However, I did not want to fix this in any regard, as you might
want to use that, and it would be a none trivial amount of code to fix
that.

Reviewed-by: Chris Michael <cp.michael@samsung.com>
Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11329

4 years agoecore_evas: introduce initial selection & dnd support for x.
Marcel Hollerbach [Sun, 12 Jan 2020 16:38:05 +0000 (17:38 +0100)]
ecore_evas: introduce initial selection & dnd support for x.

Seats are not implemented, if there is a type mismatch promises are
going to be rejected. Most of this code is copied over from
selection_manager.

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11195

4 years agoecore_x_selection: do not skip any any atoms
Marcel Hollerbach [Tue, 21 Jan 2020 13:50:23 +0000 (14:50 +0100)]
ecore_x_selection: do not skip any any atoms

i dont know why we skipped the first two atoms, but right now, if a
application is only providing one single target, we would crash.
With this we might copy a few atoms more. However, these atoms do not
matter, as we skip those, that we cannot understand

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11194

4 years agoecore_x: add API to request selection changed events for diff. wins
Marcel Hollerbach [Sun, 12 Jan 2020 16:36:48 +0000 (17:36 +0100)]
ecore_x: add API to request selection changed events for diff. wins

we need that in order to get seleciton per window events, which is
required to get a nice mapping onto the ecore_evas object.

Reviewed-by: Carsten Haitzler (Rasterman) <rasterman.com>
Reviewed-by: Chris Michael <cp.michael@samsung.com>
Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11193

4 years agoecore_evas: Introduce cnp / dnd API for ecore evas
Marcel Hollerbach [Sun, 5 Jan 2020 14:05:36 +0000 (15:05 +0100)]
ecore_evas: Introduce cnp / dnd API for ecore evas

The idea of copy and paste here is:
- The user specifies the content he wants to have in the selection
  buffer with a Eina_Content, these content pointer ownerships are
  passed to the called. Internally ecore_evas code will memorieze the
  pointer, and pass on function callbacks to the modules, which then do
  not have to deal with the ownership.

- In case the module does not specify these APIs, the callback
  implementation will be called, which only works for cnp *not* dnd.

- Action and mime types are handled as strings, which allows way better
  custom organisations.

(The docs needs improvement)

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11192

4 years agoeina: introduce Eina_Abstract_Content
Marcel Hollerbach [Sat, 4 Jan 2020 14:08:08 +0000 (15:08 +0100)]
eina: introduce Eina_Abstract_Content

A little abstraction to have abstract data content bound to a type.

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Reviewed-by: Xavi Artigas <xavierartigas@yahoo.es>
Differential Revision: https://phab.enlightenment.org/D11018

4 years agoevas_pbject_intercept: add missing parenthesis
Daniel Kolesa [Fri, 6 Mar 2020 22:30:50 +0000 (23:30 +0100)]
evas_pbject_intercept: add missing parenthesis

Broken in 7c99e0a444d81cc545181682f1a4ae00f76d0873.

*spanks*

4 years agoefl_ui_win: Fix dereference before NULL check
Christopher Michael [Fri, 6 Mar 2020 21:12:15 +0000 (16:12 -0500)]
efl_ui_win: Fix dereference before NULL check

This patch fixes an issue detected by Coverity in that 'sdp' is
already being dereferenced before we check it. ELM_WIN_DATA_GET can
return NULL, so we should check it's return Before trying to use the
variable.

Fixes CID1419871

4 years agoevas_object_intercept: Add check for NULL 'obj' variable inside macros
Christopher Michael [Fri, 6 Mar 2020 18:04:57 +0000 (13:04 -0500)]
evas_object_intercept: Add check for NULL 'obj' variable inside macros

This patch adds a small check for a valid 'obj' variable inside the
object_intercept macros, which fixes several Coverity reported issues
of NULL pointer dereference.

Fixes CID1420239 through CID1420258

4 years agoefl-wl: properly send surface enter event in smart show
Mike Blumenkrantz [Fri, 6 Mar 2020 17:16:45 +0000 (12:16 -0500)]
efl-wl: properly send surface enter event in smart show

Summary:
confusing...
Depends on D11452

Reviewers: segfaultxavi, bu5hm4n, devilhorns

Reviewed By: devilhorns

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11453

4 years agoelm: undef EAPI_WEAK in header before defining
Mike Blumenkrantz [Fri, 6 Mar 2020 17:16:40 +0000 (12:16 -0500)]
elm: undef EAPI_WEAK in header before defining

Summary: avoid double define warnings

Reviewers: devilhorns

Reviewed By: devilhorns

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11452

4 years agoevas_gl_drm: Fix incorrect define statement.
Christopher Michael [Fri, 6 Mar 2020 16:45:44 +0000 (11:45 -0500)]
evas_gl_drm: Fix incorrect define statement.

This actually should be #ifndef so that if it's not defined in the
system, then we can still compile properly.

4 years agoci: make ci work again
Marcel Hollerbach [Fri, 6 Mar 2020 12:32:53 +0000 (13:32 +0100)]
ci: make ci work again

because of some reasons, python now hides in a different PATH, which is
not *always* added to the PATH env variable.

Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org>
Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11461

4 years agoefl_canvas_layout_part: remove debugging content set / get impl
Marcel Hollerbach [Fri, 6 Mar 2020 13:52:54 +0000 (14:52 +0100)]
efl_canvas_layout_part: remove debugging content set / get impl

this is now handled when eo is erroring.

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11464

4 years agoefl_canvas_layout_part: add part name, type and group name to debug info
Marcel Hollerbach [Fri, 6 Mar 2020 13:47:52 +0000 (14:47 +0100)]
efl_canvas_layout_part: add part name, type and group name to debug info

this is usefull, as we now know which type of part we have, if a API
call does not succeed.

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11463

4 years agoeo: print debug information when a call cannot be resolved
Marcel Hollerbach [Fri, 6 Mar 2020 13:40:04 +0000 (14:40 +0100)]
eo: print debug information when a call cannot be resolved

that is usefull esp. on parts.

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11462

4 years agotests/evas: verify loading and mmap values for efl.file are correct with skip_head
Mike Blumenkrantz [Wed, 26 Feb 2020 14:42:44 +0000 (09:42 -0500)]
tests/evas: verify loading and mmap values for efl.file are correct with skip_head

ensure that these functions are still usable with async load

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D11424

4 years agoevas/image: implement efl.file loading and mmap getters to correctly handle skip_head
Mike Blumenkrantz [Wed, 26 Feb 2020 14:41:14 +0000 (09:41 -0500)]
evas/image: implement efl.file loading and mmap getters to correctly handle skip_head

when this property is set, the mixin implementation of efl_file_load() is
never called, which means the internal loaded flag (and related data) is
not set, and so the values for these properties must instead be returned
directly from the image data

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D11423

4 years agoevas: set mmap data to image data after preload with skip_head enabled
Mike Blumenkrantz [Wed, 26 Feb 2020 14:38:13 +0000 (09:38 -0500)]
evas: set mmap data to image data after preload with skip_head enabled

this is a bit ugly, but in the case where skip_head is used it's important
to propagate the resulting Eina_File back up to the image object's data
for use in other api functions

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D11422

4 years agoefl_ui/widget: add inheritance for efl.gfx.entity to widget part
Mike Blumenkrantz [Wed, 4 Mar 2020 14:47:20 +0000 (09:47 -0500)]
efl_ui/widget: add inheritance for efl.gfx.entity to widget part

most of these are empty

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D11448

4 years agoevas_object_intercept: Use efl_data_scope_safe_get
Christopher Michael [Thu, 5 Mar 2020 13:28:05 +0000 (08:28 -0500)]
evas_object_intercept: Use efl_data_scope_safe_get

As per mailing list discussion, This macro is apparently a forward
facing API so we should be using efl_data_scope_safe_get in the event
that the API receives an object of the wrong type (which would have
caused a crash previously).

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D11450

4 years agoevas_object_intercept: Reduce calls to efl_data_scope_get
Christopher Michael [Thu, 5 Mar 2020 13:24:46 +0000 (08:24 -0500)]
evas_object_intercept: Reduce calls to efl_data_scope_get

Small patch to reduce the number of calls to efl_data_scope_get as per
mailing list discussion. Since the
EVAS_OBJECT_INTERCEPT_CALLBACK_DEFINE macro already retrieves the
protected data via efl_data_scope_get, we can just pass that
protected data directly to evas_object_intercept_init/deinit functions
without the need to refetch it.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D11449

4 years agoelm_label: add EFL_ACCESS_WIDGET_ACTION_MIXIN
Marcel Hollerbach [Wed, 4 Mar 2020 14:03:20 +0000 (15:03 +0100)]
elm_label: add EFL_ACCESS_WIDGET_ACTION_MIXIN

we implement the API from it, so we should have that as a type here.

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D11447

4 years agoelm_code_syntax: Add c-sharp syntax highlighting.
Alastair Poole [Fri, 6 Mar 2020 11:03:33 +0000 (11:03 +0000)]
elm_code_syntax: Add c-sharp syntax highlighting.

4 years agoci: travis: stop osx builds until build failures are solved on Travis
Stefan Schmidt [Fri, 6 Mar 2020 11:01:46 +0000 (12:01 +0100)]
ci: travis: stop osx builds until build failures are solved on Travis

This looks like problems on the TravisCI side, but this is going on for
to long. Stop osx builds to have our on CI in a sane state again.

4 years agoelc_naviframe: fix to delete item once by elm_object_item_del in pop_cb
Jaehyun Cho [Fri, 6 Mar 2020 04:25:04 +0000 (13:25 +0900)]
elc_naviframe: fix to delete item once by elm_object_item_del in pop_cb

If elm_object_item_del is called in pop_cb and pop_cb returns
EINA_FALSE, then the given item is destructed by _item_noref when
efl_unref is called after pop_cb.

After the above destruction, efl_del is called after the above efl_unref
and it deletes the item again.

Not to delete item after the item is destructed, efl_del after pop_cb is
removed.

4 years agoelm_code: Avoid moving to non-existent line.
Alastair Poole [Fri, 6 Mar 2020 01:23:35 +0000 (01:23 +0000)]
elm_code: Avoid moving to non-existent line.

Check for line existence before moving cursor.

4 years agoevas: Fix dereferenceing null pointer
Christopher Michael [Wed, 4 Mar 2020 19:20:47 +0000 (14:20 -0500)]
evas: Fix dereferenceing null pointer

Coverity reports a forward null here. If we do not get obj from
efl_data_scope_get then we should return. Also fixes some formatting

Fixes CID1420216

4 years agoevas: Fix mouse event info for legacy events.
Hosang Kim [Wed, 4 Mar 2020 14:35:30 +0000 (09:35 -0500)]
evas: Fix mouse event info for legacy events.

Summary:
Legacy event info have canvas and output coordinates.
Output coordinates have information of original position.
Canvas coordinates have information of transformed position.
So keep backward compatibility, fix filling legacy information.

This reverts commit 7f724f6c5db1127a6dd5786791f495b58910e9a9

Reviewers: Hermet, zmike

Reviewed By: zmike

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11445

4 years agoelm_code: Improve efficiency of the widget.
Alastair Poole [Wed, 4 Mar 2020 01:41:43 +0000 (01:41 +0000)]
elm_code: Improve efficiency of the widget.

For now, only create textgrids when needed. Also improve other
regions of the widget. This will improve large file support and
also some rendering of the widget.

This improves performance 1-2 times. However, there must be a
reasonable redesign regarding the current use of textgrids.

4 years agoedje: Change the order of Edje_Action_Type enum members 82/227082/1
JunsuChoi [Mon, 9 Mar 2020 11:01:33 +0000 (20:01 +0900)]
edje: Change the order of Edje_Action_Type enum members

The tizen version of tizen studio has not been updated yet,
so we should keep EDJE_ACTION_TYPE_RUN_PLUGIN equal to 25.

@tizen_only

Change-Id: I4233471f830569b04c1df1723d7e06013131e1df

4 years agoEcore_Wl2: Fix doc. 64/227064/2
Woochanlee [Mon, 9 Mar 2020 09:31:45 +0000 (18:31 +0900)]
Ecore_Wl2: Fix doc.

Add @brief.
Modify some of the doc.
Delete wrong indexing.

Change-Id: I6fd79fbaa3603977fbe1b87484f8f7939f02fae1

4 years agoEfl.Ui.Image: Modify current frame to start at 1 61/227061/1
JunsuChoi [Mon, 9 Mar 2020 08:54:33 +0000 (17:54 +0900)]
Efl.Ui.Image: Modify current frame to start at 1

Summary:
3dc3deb patch was changed to start at 1 rather than 0.
gif image starts at 1, not 0.

Test Plan: N/A

Reviewers: Hermet, kimcinoo

Reviewed By: Hermet

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11474

Change-Id: I28140bcf343b24a039f3b83fccead03099c3edc4

4 years agoEcore_Wl2: Introduce ecore_wl2_window_native_surface_get API. 25/227025/2 submit/tizen/20200309.064055
Woochanlee [Mon, 9 Mar 2020 06:18:50 +0000 (15:18 +0900)]
Ecore_Wl2: Introduce ecore_wl2_window_native_surface_get API.

This API to support some apps may using native surface handle.

@tizen_only

Change-Id: Ica07a97dfe75d226e8a6dbec2dd7230002dadc00

4 years agoefl_wl: Prevent a null pointer dereference. 14/227014/1
Woochanlee [Mon, 9 Mar 2020 05:34:56 +0000 (14:34 +0900)]
efl_wl: Prevent a null pointer dereference.

Change-Id: Ie11cdc246e2551864ca8ab6e0ab7cba3b9bef96c
CID:1133579

4 years agoelc_naviframe: fix to delete item once by elm_object_item_del in pop_cb 00/227000/1
Jaehyun Cho [Fri, 6 Mar 2020 04:25:04 +0000 (13:25 +0900)]
elc_naviframe: fix to delete item once by elm_object_item_del in pop_cb

If elm_object_item_del is called in pop_cb and pop_cb returns
EINA_FALSE, then the given item is destructed by _item_noref when
efl_unref is called after pop_cb.

After the above destruction, efl_del is called after the above efl_unref
and it deletes the item again.

Not to delete item after the item is destructed, efl_del after pop_cb is
removed.

Change-Id: Icc3716e543d8f017f054bac62c2b24809b76e8da

4 years agoecore_ipc: remove efl_loop_fd_set before finishing the unix sockect connection 99/226799/2
Yeongjong Lee [Thu, 5 Mar 2020 10:03:17 +0000 (19:03 +0900)]
ecore_ipc: remove efl_loop_fd_set before finishing the unix sockect connection

For Efl.Net.Dialer_Unix, fd should be set when the socket is finished connecting.

efl_loop_fd_set is called in the `_efl_net_dialer_unix_connected` which is
called at the end of connection thread named `_efl_net_connect_async_end`.
That means we don't need to set fd right after `efl_add(EFL_NET_DIALER_UNIX_CLASS`.

Also, this will fix intermittent efl-test-suite failure related to ecore_evas_extn.

@tizen_fix

Change-Id: I2f884e3de49be109cc945c76f42d22c206e10c1b

4 years agoManual cherry-pick from here, cause of fetch network issue...
Hermet Park [Sun, 8 Mar 2020 11:42:16 +0000 (20:42 +0900)]
Manual cherry-pick from here, cause of fetch network issue...

commit b2bd0c548f03b8178e40e0cfa6e9c31c4bff38b1 (HEAD -> master, origin/master, origin/HEAD)
Author: Hermet Park <chuneon.park@samsung.com>
Date:   Sun Mar 8 20:20:43 2020 +0900

    evas gl: clip rect with proper size.

    gc width/height indicates the main output size,
    if the target surface(fbo) is larger than this, rect won't be drawn.

    We should clip this with current context size.

    @fix

Change-Id: Ida057ce7561b8cfe60b5d3339a7e095bf1a1b94c

4 years agocanvas grid: fix inifinite loop dead.
Hermet Park [Fri, 6 Mar 2020 14:48:43 +0000 (23:48 +0900)]
canvas grid: fix inifinite loop dead.

Here parent smart del() call occurs inifinite loop for member deletion,

Maybe we need a better clean and neat solution here.

Change-Id: I67155aa916243c03306cd82eeb7d58fcb4491703

4 years agoelm/atspi_bridge: fix memory leak 11/226811/5
Kamil Konieczny [Thu, 5 Mar 2020 10:50:23 +0000 (11:50 +0100)]
elm/atspi_bridge: fix memory leak

Fix memory leak in function elm_atspi_bridge_utils_say,
as function _elm_util_mkup_to_text return allocated memory.
This also fixes functional bug in text markup removal from
original text. Also while at this, add error checks.

Change-Id: Ife37a0ae66caf677174bef7bd7f5bc9b633b9357

4 years agoFix the bug of eglSync for EvasGL 45/226645/2 submit/tizen/20200306.082439
Wonsik Jung [Wed, 4 Mar 2020 09:10:01 +0000 (18:10 +0900)]
Fix the bug of eglSync for EvasGL

eglWaitSync is called with the destoried eglSync handle, sometimes.
This patch is for the case.

Change-Id: Iad2f7d5332c4b635d98f7ee72152b4accb773a3c

4 years agoelm_genlist: check viewport height to calculate for item show 01/226801/3 submit/tizen/20200305.112415
Jaehyun Cho [Thu, 5 Mar 2020 09:30:44 +0000 (18:30 +0900)]
elm_genlist: check viewport height to calculate for item show

To show item in the correct position, viewport height should be
calculated correctly.

Since the default height of viewport is 1, correctly calculated viewport
height is bigger than 1.

Therefore, viewport height is checked if it is bigger than 1 to
calculate for item show.

Change-Id: If78db939fa2fb0bcca8be9ffe7e23c297c7852a0

4 years agoRevert "elm/image: restore legacy behavior of elm_image_file_get" 90/226790/3
Bowon Ryu [Thu, 5 Mar 2020 09:38:03 +0000 (18:38 +0900)]
Revert "elm/image: restore legacy behavior of elm_image_file_get"

This reverts commit dec742a9357050bcd29a9eed0e9e49320d9c146d.

This patch returns an absolute path even if user file_set is a relative path.
This causes compatibility issues.

Change-Id: If0304569418f9f00c344d5e8a4a8ae373fc163a8

4 years agoelementary: use array instead of list for subchildren 15/226615/2
Marcel Hollerbach [Mon, 17 Feb 2020 12:56:06 +0000 (13:56 +0100)]
elementary: use array instead of list for subchildren

this safes in the runtime of elm_test round about 9kb (*).
Additionally, using a array here is jumping way fewer times arround in
memory, as we do not need to jump from node to node in a list.

Additionally, this commit replaces a few abort disabler macros with a
error checking macro. (which cleans the log).

*: explanation: we have round about 600 widgets in elm_test, every
widget is normally refered once, every list node has 4 pointer, makes
round about 9600 bytes or rougly 9 KB. So the messured savings are more
or less explaining the reality.

Reviewed-by: Carsten Haitzler (Rasterman) <rasterman.com>
Differential Revision: https://phab.enlightenment.org/D11374

Change-Id: Ib906697a1ba71277e1fbdebedbdc0c05cb44d53a

4 years agoelementary: Remove internal flag of Elm_Win_Aux_Message structure 68/226768/1
Doyoun Kang [Thu, 5 Mar 2020 07:36:17 +0000 (16:36 +0900)]
elementary: Remove internal flag of Elm_Win_Aux_Message structure

@tizen_only

Change-Id: I25884f7e67dbfbebb576dc1af392a9ba3c04210c
Signed-off-by: Doyoun Kang <doyoun.kang@samsung.com>
4 years agoEcore_Wl2: Add dox and rearrange API list. 60/226660/2
Woochanlee [Wed, 4 Mar 2020 11:16:51 +0000 (20:16 +0900)]
Ecore_Wl2: Add dox and rearrange API list.

Change-Id: Ief34bd68508c03254ff527d171cf6e811d145548

4 years agoRemove internal flag of elm_win_aux_msg APIs 52/226552/2
Doyoun Kang [Tue, 3 Mar 2020 05:42:53 +0000 (14:42 +0900)]
Remove internal flag of elm_win_aux_msg APIs

- elm_win_aux_msg_key_get
- elm_win_aux_msg_val_get
- elm_win_aux_msg_options_get

@tizen_only

Change-Id: I61fcc31753d02cde1cc30c60cfe169e54f9cf027
Signed-off-by: Doyoun Kang <doyoun.kang@samsung.com>
4 years agoevas_object_main: add checking delete_me when submit/tizen/20200303.214157
Hosang Kim [Tue, 3 Mar 2020 09:58:02 +0000 (18:58 +0900)]
evas_object_main: add checking delete_me when
_evas_object_pointer_data_get is called.

Sometimes Eo object is already deleted, but
_evas_object_pointer_data_get is called.
So add checking delete_me for protecting crash.

Change-Id: I682a7095df1744745ce29e2af8d7be04781aaf4e

4 years agogenlist: refactoring item deletion process
SangHyeon Jade Lee [Mon, 2 Mar 2020 09:35:09 +0000 (18:35 +0900)]
genlist: refactoring item deletion process

sd->items must be cleaned before some item deletion process,
so that adjust item cannot find deleted item on sd->items list.

Change-Id: I3f862cc9d5b39291d7febcfcdfb4efbfe3937753

4 years ago[access] Handle case where last highlighted object is in different window
Prasoon Singh [Wed, 27 Nov 2019 11:17:16 +0000 (16:47 +0530)]
[access] Handle case where last highlighted object is in different window

When highlight_grab is called simultaneously fo different window object
highlight frame remains on last window.

Change-Id: I60dbfefda83a68aaa641aa289e92d3b99f6add26
Signed-off-by: Prasoon Singh <prasoon.16@samsung.com>
4 years agoecore_timer: keep backward compatibility on freeze/thaw behavior
Hosang Kim [Tue, 25 Feb 2020 05:54:27 +0000 (14:54 +0900)]
ecore_timer: keep backward compatibility on freeze/thaw behavior

Legacy timer is thawed when ecore_timer_thaw is called even if freeze
and thaw pairs don't match.

Change-Id: I33bcb973a3f1d990e212cd09bdc36530bfcfbe19

4 years agoecore_wl2_input: fix repeat key window.
Hosang Kim [Mon, 24 Feb 2020 12:00:10 +0000 (21:00 +0900)]
ecore_wl2_input: fix repeat key window.

when window is removing, key repeat timer should be removed also.

Change-Id: I8971aa04b57379d89f117b4b9b4332a33561436c

4 years agoedje_message_queue: add checking number of messages.
Hosang Kim [Fri, 21 Feb 2020 09:51:37 +0000 (18:51 +0900)]
edje_message_queue: add checking number of messages.

Sometimes edje_message_del is called while processing edje message.
It makes double free corruption.

Change-Id: I06f01a8670d4220b20503790990ff8d751e47ff2

4 years agoedje_message_queue: decrease number of messages when message is removed.
Hosang Kim [Fri, 21 Feb 2020 07:02:07 +0000 (16:02 +0900)]
edje_message_queue: decrease number of messages when message is removed.

Reviewers: raster

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11393

Change-Id: Ia7ba6de2061432bd503557a987e073504d4e3664
Signed-off-by: Hosang Kim <hosang12.kim@samsung.com>
4 years agoRevert "elementary: use array instead of list for subchildren"
Jongmin Lee [Thu, 20 Feb 2020 22:06:01 +0000 (07:06 +0900)]
Revert "elementary: use array instead of list for subchildren"

This reverts commit cf583498a72be5f18b2dd6a4fb10ff1f1dbca53f.

4 years agoelementary: use array instead of list for subchildren
Marcel Hollerbach [Mon, 17 Feb 2020 12:56:06 +0000 (13:56 +0100)]
elementary: use array instead of list for subchildren

this safes in the runtime of elm_test round about 9kb (*).
Additionally, using a array here is jumping way fewer times arround in
memory, as we do not need to jump from node to node in a list.

Additionally, this commit replaces a few abort disabler macros with a
error checking macro. (which cleans the log).

*: explanation: we have round about 600 widgets in elm_test, every
widget is normally refered once, every list node has 4 pointer, makes
round about 9600 bytes or rougly 9 KB. So the messured savings are more
or less explaining the reality.

Reviewed-by: Carsten Haitzler (Rasterman) <rasterman.com>
Differential Revision: https://phab.enlightenment.org/D11374

Change-Id: I8653d875ab8bd7db4a48910dc2ed08cee1e76623

4 years agoavahi - let's disable by default as there doesn't seem to be a use
Carsten Haitzler (Rasterman) [Mon, 2 Mar 2020 19:02:09 +0000 (19:02 +0000)]
avahi - let's disable by default as there doesn't seem to be a use

i asked why we should have it by default etc. and what it's needed
for. i cant't find any... and no answer soc off by default to trim efl
down.

4 years agoefl-wl: add toplevel_added event for adding a new toplevel surface
Mike Blumenkrantz [Tue, 3 Mar 2020 18:42:47 +0000 (13:42 -0500)]
efl-wl: add toplevel_added event for adding a new toplevel surface

Summary: Depends on D11443

Reviewers: devilhorns

Reviewed By: devilhorns

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11444

4 years agoefl-wl: ensure that child_added event is called only when a parent is set
Mike Blumenkrantz [Tue, 3 Mar 2020 18:42:43 +0000 (13:42 -0500)]
efl-wl: ensure that child_added event is called only when a parent is set

Reviewers: devilhorns

Reviewed By: devilhorns

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11443

4 years agodoxygen docs: Fix several group problems
Xavi Artigas [Tue, 3 Mar 2020 17:46:36 +0000 (18:46 +0100)]
doxygen docs: Fix several group problems

@ingroup before @addgroup, circular group dependencies...

4 years agodoxygen docs: fix insufficiently namespaced group names
Xavi Artigas [Tue, 3 Mar 2020 16:49:23 +0000 (17:49 +0100)]
doxygen docs: fix insufficiently namespaced group names

These led to clashes, since group names must be globally unique.

4 years agodoxygen docs: fix multiple defined @section's
Xavi Artigas [Tue, 3 Mar 2020 10:09:47 +0000 (11:09 +0100)]
doxygen docs: fix multiple defined @section's

Section names must be globally unique.
We rarely reference sections so we don't actually need to use @section
and a simple header suffices.

4 years agoefl-wl: add function to get pid of a surface
Mike Blumenkrantz [Tue, 3 Mar 2020 17:27:43 +0000 (12:27 -0500)]
efl-wl: add function to get pid of a surface

Summary: this is useful for tracking surfaces of specific clients

Reviewers: bu5hm4n, devilhorns

Reviewed By: devilhorns

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D11438

4 years agotests_: elementary: remove not used allocation in collection view test
Stefan Schmidt [Mon, 2 Mar 2020 15:46:19 +0000 (16:46 +0100)]
tests_: elementary: remove not used allocation in collection view test

Allocated but never used (and never freed). Seems like a copy and paste
bug to me.

CID: 1412363

Signed-off-by: Stefan Schmidt <s.schmidt@samsung.com>
Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D11442

4 years agotests_: elementary: free memory in error path
Stefan Schmidt [Mon, 2 Mar 2020 15:40:38 +0000 (16:40 +0100)]
tests_: elementary: free memory in error path

Make sure we free the iterator here as well.

CID: 1409658

Signed-off-by: Stefan Schmidt <s.schmidt@samsung.com>
Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D11441

4 years agotests/gesture: add test for gesture sequence
Mike Blumenkrantz [Mon, 2 Mar 2020 15:23:31 +0000 (10:23 -0500)]
tests/gesture: add test for gesture sequence

ensure that tap continues working after complex gestures

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D11440

4 years agoefl/gesture: deduplicate some code
Mike Blumenkrantz [Thu, 20 Feb 2020 18:55:40 +0000 (13:55 -0500)]
efl/gesture: deduplicate some code

also remove a misleading comment

no functional changes

Differential Revision: https://phab.enlightenment.org/D11395

4 years agotests/gesture: add a couple rotate gesture tests
Mike Blumenkrantz [Thu, 20 Feb 2020 18:48:14 +0000 (13:48 -0500)]
tests/gesture: add a couple rotate gesture tests

these end up triggering a lot of corner cases in other recognizers too,
but the tests themselves are fairly minimal

Differential Revision: https://phab.enlightenment.org/D11391

4 years agotests/gesture: verify rotate gesture event counting in existing tests
Mike Blumenkrantz [Wed, 19 Feb 2020 17:57:26 +0000 (12:57 -0500)]
tests/gesture: verify rotate gesture event counting in existing tests

these magically pass already

Differential Revision: https://phab.enlightenment.org/D11390

4 years agotests/elm: add util function for doing a multi-press drag around
Mike Blumenkrantz [Thu, 20 Feb 2020 18:24:52 +0000 (13:24 -0500)]
tests/elm: add util function for doing a multi-press drag around

same as existing function, but takes a finger id

Differential Revision: https://phab.enlightenment.org/D11389