platform/upstream/efl.git
6 years agoelc_multibuttonentry: Fix to make multibuttonentry to be legacy only
Jaehyun Cho [Fri, 20 Apr 2018 02:48:19 +0000 (11:48 +0900)]
elc_multibuttonentry: Fix to make multibuttonentry to be legacy only

To make multibuttonentry to be legacy only, inappropriate efl ui
interface usages are removed in multibuttonentry.

Change-Id: Id73ae1302447823129e9eb9c84c44bc4c373cd23

6 years agoecore: fix a potential memory leak.
Hermet Park [Fri, 20 Apr 2018 04:29:46 +0000 (13:29 +0900)]
ecore: fix a potential memory leak.

if the func argument is null, allocated memory won't be freed properly.
return it as soon as possible.

@fix

6 years agoefl_mono: Fix after prop get const change
Lauro Moura [Thu, 19 Apr 2018 17:01:49 +0000 (14:01 -0300)]
efl_mono: Fix after prop get const change

6 years agoFocus: If tree focus is set to not allow focus on the object - don't give it focus.
Stephen 'Okra' Houston [Thu, 19 Apr 2018 16:17:34 +0000 (11:17 -0500)]
Focus: If tree focus is set to not allow focus on the object - don't give it focus.

Patch from Marcel 'bu5hm4n' Hollerbach

6 years agoelementary: Fix multibuttonentry legacy build.
Lauro Moura [Thu, 19 Apr 2018 15:09:37 +0000 (12:09 -0300)]
elementary: Fix multibuttonentry legacy build.

Summary:
The legacy header was still referencing the eo class removed in
939690a8a99.

Reviewers: cedric, Jaehyun_Cho, vitor.sousa

Reviewed By: vitor.sousa

Differential Revision: https://phab.enlightenment.org/D5970

6 years agotest_ui_stack: Add test using Efl.Ui.Stack class
Jaehyun Cho [Thu, 12 Apr 2018 12:45:32 +0000 (21:45 +0900)]
test_ui_stack: Add test using Efl.Ui.Stack class

Add test using Efl.Ui.Stack, Efl.Ui.Navigation_Layout, and
Efl.Ui.Navigation_Bar classes.

Navigation_Bar widget is set into Navigation_Layout widget and
Navigation_Layout widget is pushed into Stack widget.

6 years agoefl_ui_navigation_bar: Add Efl.Ui.Navigation_Bar class
Jaehyun Cho [Thu, 12 Apr 2018 12:42:38 +0000 (21:42 +0900)]
efl_ui_navigation_bar: Add Efl.Ui.Navigation_Bar class

Efl.Ui.Navigation_Bar is a widget which provides a bar form useful for
navigation.

6 years agoefl_ui_navigation_layout: Add Efl.Ui.Navigation_Layout class
Jaehyun Cho [Thu, 12 Apr 2018 12:30:05 +0000 (21:30 +0900)]
efl_ui_navigation_layout: Add Efl.Ui.Navigation_Layout class

Efl.Ui.Navigation_Layout is a widget which provides a layout form useful
for navigation.

6 years agoefl_ui_stack: Add Efl.Ui.Stack class
Jaehyun Cho [Thu, 12 Apr 2018 12:09:38 +0000 (21:09 +0900)]
efl_ui_stack: Add Efl.Ui.Stack class

Efl.Ui.Stack is a container arranges objects in stack structure by
pushing and popping them.

6 years agoefl_canvas_object: Add events for state changes of object's animation
Jaehyun Cho [Tue, 10 Apr 2018 13:09:39 +0000 (22:09 +0900)]
efl_canvas_object: Add events for state changes of object's animation

Add events to notice the state changes of canvas object's animation.
anim_started, anim_running, and anim_ended events are added.

6 years agoRevert Efl.Ui.Multibuttonentry to create new one.
Woochan Lee [Thu, 19 Apr 2018 13:20:29 +0000 (22:20 +0900)]
Revert Efl.Ui.Multibuttonentry to create new one.

Summary:
Revert e02b2f04c2945ad60fab5612af1e02b0838b7ff5.

I couldnt make a revert commit for many commit has been related with above commit.

I will create a new MBE eo class ASAP. (https://phab.enlightenment.org/T5358)

Reviewers: cedric, woohyun, Jaehyun_Cho, SanghyeonLee, herb

Reviewed By: Jaehyun_Cho

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5954

6 years agoeolian: re-enable non-dependencied parsing
Daniel Kolesa [Thu, 19 Apr 2018 11:40:41 +0000 (13:40 +0200)]
eolian: re-enable non-dependencied parsing

The issue was that standalone-parsed files (outside of dependencies
of some other unit) were not being included in that dependency tree
(obviously) which resulted in their own dependency tree being
excluded from merging, causing the database hashes to lack the
necessary elements.

To fix this, I trigger merging of the "standalone" dependency trees
separately, by calling _merge_units from defer parsing cb.

6 years agoeolian: disable non-dependencied parsing for the time being
Daniel Kolesa [Wed, 18 Apr 2018 19:16:04 +0000 (21:16 +0200)]
eolian: disable non-dependencied parsing for the time being

As 543902f2ba13b69b052eb21a4a36324d96f5eeba breaks build,
disable parsing with state as a parent for now, until the issue
is figured out.

6 years agoelm: Move legacy genlist types to elm_general.
Lauro Moura [Wed, 18 Apr 2018 17:03:11 +0000 (19:03 +0200)]
elm: Move legacy genlist types to elm_general.

Summary: Fixes T6771

Reviewers: bu5hm4n, felipealmeida

Subscribers: q66, cedric

Maniphest Tasks: T6771

Differential Revision: https://phab.enlightenment.org/D5938

6 years agoeolian: make doc ref resolution global
Daniel Kolesa [Wed, 18 Apr 2018 14:34:49 +0000 (16:34 +0200)]
eolian: make doc ref resolution global

Doc refs no longer introduce new dependencies into files. Instead,
they're parsed globally, and any doc ref lookup is also made
globally. This allows unit based dependencies to correspond more
to what files actually really need at compile time/runtime, with
docs being irrelevant to that; it also simplifies the API.

The doc resolution API now takes Eolian_State instead of
Eolian_Unit, too.

6 years agoeolian: doc ref validation cleanup
Daniel Kolesa [Wed, 18 Apr 2018 13:26:11 +0000 (15:26 +0200)]
eolian: doc ref validation cleanup

6 years agoeolian: add internal api for dep deferring + parsing without dep
Daniel Kolesa [Wed, 18 Apr 2018 13:23:07 +0000 (15:23 +0200)]
eolian: add internal api for dep deferring + parsing without dep

This is cleaner than adding into a hash manually. Additionally, it
is now possible to request that the file be parsed not as a dep,
but rather standalone, which will be useful later.

6 years agoeolian: simpler and more efficient defer parsing
Daniel Kolesa [Wed, 18 Apr 2018 13:04:04 +0000 (15:04 +0200)]
eolian: simpler and more efficient defer parsing

6 years agoelm_win: Don't do auto norender on non X11 engines
Derek Foreman [Tue, 17 Apr 2018 19:01:39 +0000 (14:01 -0500)]
elm_win: Don't do auto norender on non X11 engines

This only really makes sense on X11 and can lead to some seriously
confusing cases on other engines (*cough* wayland) when elm's idea
of iconified state doesn't match the compositor's.

While currently only X11 is whitelisted, other backends can be
added, though I suspect most are more like wayland where it makes
no sense at all.

ref T6834

6 years agotests: preload elm csd theme in elm_suite startup
Mike Blumenkrantz [Wed, 18 Apr 2018 11:58:15 +0000 (13:58 +0200)]
tests: preload elm csd theme in elm_suite startup

Summary:
ref T6864
Depends on D5958

Reviewers: stefan_schmidt

Subscribers: cedric

Maniphest Tasks: T6864

Differential Revision: https://phab.enlightenment.org/D5959

6 years agotests: skip second ecore_shutdown call in elm_suite when forking
Mike Blumenkrantz [Wed, 18 Apr 2018 11:57:39 +0000 (13:57 +0200)]
tests: skip second ecore_shutdown call in elm_suite when forking

Summary:
init is only called here to verify the refcount, shutdown is only necessary
when not forking in order to preserve the environment

ref T6864

Reviewers: stefan_schmidt

Subscribers: cedric

Maniphest Tasks: T6864

Differential Revision: https://phab.enlightenment.org/D5958

6 years agoexamples: cxx: fix some missing const attributes in colourable*
Stefan Schmidt [Wed, 18 Apr 2018 08:11:43 +0000 (10:11 +0200)]
examples: cxx: fix some missing const attributes in colourable*

Fixing some fallout from fcae7cab276c2d3b19bb913ad908b67b7dd33089

Given that the examples do not build for weeks with CXX enabled this is
just the tip of the iceberg.

6 years agotravis: fix install of packages on macOS
Stefan Schmidt [Wed, 18 Apr 2018 07:24:18 +0000 (09:24 +0200)]
travis: fix install of packages on macOS

Some recent OSX builds failed on Travis due to a change on the macOS
image they are providing. The python package could not be installed.

==> Pouring python@2-2.7.14_3.sierra.bottle.1.tar.gz
Error: The `brew link` step did not complete successfully
The formula built, but is not symlinked into /usr/local
Could not symlink bin/2to3-2
Target /usr/local/bin/2to3-2
is a symlink belonging to python. You can unlink it:
  brew unlink python

6 years agoconfig: fix Efl.Ui.Slider key binding
YeongJong Lee [Wed, 18 Apr 2018 06:34:35 +0000 (15:34 +0900)]
config: fix Efl.Ui.Slider key binding

Summary: this fixes T6796

Test Plan: elemetary_test -to slider

Reviewers: woohyun, Hermet, Jaehyun_Cho

Reviewed By: Jaehyun_Cho

Subscribers: cedric

Maniphest Tasks: T6796

Differential Revision: https://phab.enlightenment.org/D5962

6 years agoecore_con: Fix build
Derek Foreman [Tue, 17 Apr 2018 19:49:20 +0000 (14:49 -0500)]
ecore_con: Fix build

commit fcae7cab276c2d3b19bb913ad908b67b7dd33089 broke connman stuff,
I've blindly fixed it up here.  I don't actually use or test ant of this
but apparently I build it.

If it's broken, please mail the pieces to Daniel Kolesa.

Ref T6859 I guess

6 years agoeolian gen: enable constness generation on property getter impls
Daniel Kolesa [Tue, 17 Apr 2018 18:09:44 +0000 (20:09 +0200)]
eolian gen: enable constness generation on property getter impls

This changes a lot of things all across the EFL. Previously,
methods tagged @const had both their external prototype and
internal impl generated with const on object, while property
getters only had const on the external API. This is now changed
and it all has const everywhere.

Ref T6859.

Change-Id: I853d8a8ea9fc082e75c1932e570e6693dfb3c24c

6 years agoecore_wl2: handle configure_complete after handling configure
Derek Foreman [Tue, 17 Apr 2018 16:12:27 +0000 (11:12 -0500)]
ecore_wl2: handle configure_complete after handling configure

since configure_complete may result in a draw op, configure should be
handled first so the size is correct.

6 years agotests: preload elm_init and default theme groups in elm_suite
Mike Blumenkrantz [Tue, 17 Apr 2018 13:43:52 +0000 (15:43 +0200)]
tests: preload elm_init and default theme groups in elm_suite

Summary:
opening the default theme file can take up to 0.2s in worst case during
a test

ref T6864
Depends on D5914

Reviewers: stefan_schmidt

Reviewed By: stefan_schmidt

Subscribers: cedric

Maniphest Tasks: T6864

Differential Revision: https://phab.enlightenment.org/D5915

6 years agotests: use loop test to split jpeg tests from evas_object_image_all_loader_data
Mike Blumenkrantz [Tue, 17 Apr 2018 09:55:20 +0000 (11:55 +0200)]
tests: use loop test to split jpeg tests from evas_object_image_all_loader_data

Summary:
jpeg loader tests take 4+ seconds due to manual value comparisons, so ensure
that they run separately

ref T6856

Reviewers: stefan_schmidt

Subscribers: cedric

Maniphest Tasks: T6856

Differential Revision: https://phab.enlightenment.org/D5911

6 years agoelm_dbus_menu: add missing const for Eina_List
YeongJong Lee [Tue, 17 Apr 2018 02:08:25 +0000 (11:08 +0900)]
elm_dbus_menu: add missing const for Eina_List

Summary:
This fixes following warning

../src/lib/eina/eina_list.h:1421:10: warning: assignment discards â€˜const’ qualifier from pointer target type [-Wdiscarded-qualifiers]

Test Plan: make

Reviewers: SanghyeonLee

Reviewed By: SanghyeonLee

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5952

6 years agoRevert "Genlist: restore old behaviour of item next/prev in group items"
Dave Andreoli [Mon, 16 Apr 2018 22:11:11 +0000 (00:11 +0200)]
Revert "Genlist: restore old behaviour of item next/prev in group items"

This reverts commit f0a0da9f449b0878fe6a5ce2abc50b8b6589c50a.

As per T5938 seems we really want to restore a totally wrong
behaviour, without taking care of newer apps being broken.

I revert this for now, but I'm still convinced that we must
find a way to let user use a sane ordering for newer app.

I'm thinking about adding an api in genlist to let the widget
use the new sane ordering, something like
elm_genlist_fixed_ordering_set(bool) so that new apps can use
this to ensure correct behaviour. zmike what do you think about
this solution?

6 years agoedje: Dont append font source if fonts are not embedded in edje
SubodhKumar [Mon, 16 Apr 2018 18:23:21 +0000 (11:23 -0700)]
edje: Dont append font source if fonts are not embedded in edje

Summary:
If font source is not embedded in edje, it should not
be appended in font source to skip the lookup in edj file
while loading font.

@fix

Test Plan: NA

Reviewers: cedric

Reviewed By: cedric

Subscribers: shilpasingh

Differential Revision: https://phab.enlightenment.org/D5909

Reviewed-by: Cedric BAIL <cedric@osg.samsung.com>
6 years agoeina: Improve documentation formatting
Bryce Harrington [Mon, 16 Apr 2018 18:21:48 +0000 (11:21 -0700)]
eina: Improve documentation formatting

Summary:
Revise wording of a few parameters, improve consistency of punctuation
and documentation body placement.

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5918

Reviewed-by: Cedric BAIL <cedric@osg.samsung.com>
6 years agoeina: Cleanup documentation for Sparse Matrix.
Bryce Harrington [Mon, 16 Apr 2018 18:20:21 +0000 (11:20 -0700)]
eina: Cleanup documentation for Sparse Matrix.

Summary:
Revises some descriptions, polishes wording, makes punctuation and
formatting more consistent, and straightens out some convoluted grammar
here and there.

There are a few places where the parameter documentation was incorrectly
cut-and-pasted from other routines, so this is fixed.  Also, in
eina_matrixsparse_iterator_new() the docs say the elements are not
returned sequentially, but from the context and code I think it meant to
say they're not _consecutive_.

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5924

Reviewed-by: Cedric BAIL <cedric@osg.samsung.com>
6 years agoFix a shutdown logic to be reachable in some case
Wonki Kim [Mon, 16 Apr 2018 18:18:27 +0000 (11:18 -0700)]
Fix a shutdown logic to be reachable in some case

Summary:
move evas_async_events_shutdown: to out of EVAS_CSERVE2 ifdef block
to make it reachable.

Reviewers: cedric, woohyun

Differential Revision: https://phab.enlightenment.org/D5926

Reviewed-by: Cedric BAIL <cedric@osg.samsung.com>
6 years agoelm_menu, toolbar, ctxpopup : remove unnecesary interface efl_ui_item and efl_ui_menu
SangHyeon Lee [Mon, 16 Apr 2018 18:02:59 +0000 (11:02 -0700)]
elm_menu, toolbar, ctxpopup : remove unnecesary interface efl_ui_item and efl_ui_menu

Summary:
Legacy menu toolbar and ctxpopup are not new efl interface widget,
but it have efl interfaces, efl_ui_item and efl_ui_menu.
if we plan to implement this two, it should be a class not a interface,
so I remove it on legacy widget.

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5939

Change-Id: I7356906482fd357eed7d7e576d08277375dd2e15
Reviewed-by: Cedric BAIL <cedric@osg.samsung.com>
6 years agoecore: further reduce pipe wait time during shutdown.
Mike Blumenkrantz [Mon, 16 Apr 2018 17:53:21 +0000 (10:53 -0700)]
ecore: further reduce pipe wait time during shutdown.

Summary:
it turns out that 0.01s is actually a lot, accounting for something like
20s across a run of 'make check' while providing no additional value

ref e0c8ab4c792c3ded7f4189a890a1cc7ab18e76e0
ref T6825
ref T6864

Reviewers: cedric

Maniphest Tasks: T6864, T6825

Differential Revision: https://phab.enlightenment.org/D5941

Reviewed-by: Cedric BAIL <cedric@osg.samsung.com>
6 years agotests: remove explicit checks for eina mempool tests
Mike Blumenkrantz [Mon, 16 Apr 2018 09:46:12 +0000 (11:46 +0200)]
tests: remove explicit checks for eina mempool tests

Summary:
this was only a temporary measure
Depends on D5916

Reviewers: stefan_schmidt

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5937

6 years agotests: flush stdout after printing test suite timing info
Mike Blumenkrantz [Mon, 16 Apr 2018 09:46:01 +0000 (11:46 +0200)]
tests: flush stdout after printing test suite timing info

Summary:
avoid repeating lines

Depends on D5937

Reviewers: stefan_schmidt

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5917

6 years agotests: do not attempt to parallelize test suites with only 1 test case
Mike Blumenkrantz [Mon, 16 Apr 2018 09:45:49 +0000 (11:45 +0200)]
tests: do not attempt to parallelize test suites with only 1 test case

Summary: Depends on D5911

Reviewers: stefan_schmidt

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5916

6 years agoelm_list: check there is focused item when list get focus at first
YeongJong Lee [Sun, 15 Apr 2018 09:08:14 +0000 (11:08 +0200)]
elm_list: check there is focused item when list get focus at first

Summary:
list have got focus. but there isn't logic to check whether focused_item exist.
we need that logic for list focus
this fixes T6807, T6799

Test Plan: elementary_test -to 'list focus'

Reviewers: bu5hm4n

Reviewed By: bu5hm4n

Subscribers: cedric

Maniphest Tasks: T6807, T6799

Differential Revision: https://phab.enlightenment.org/D5935

6 years agoGenlist: restore old behaviour of item next/prev in group items
Dave Andreoli [Fri, 13 Apr 2018 18:37:36 +0000 (20:37 +0200)]
Genlist: restore old behaviour of item next/prev in group items

Commit fd82c2521 has changed the behaviour of item_next/prev_get()
in case of genlist with group items (not tree).

The lagacy behaviour was to tread normal items and group items
in a flat way, like if group items was on the same level of
the children normal items.

As the commit already implement bug compatibility, seems to me
the case to restore also this case. Note that this changes only
apply to legacy genlist (I think).

Let me know if this broke something for you...as touching genlist
code is always an "I hope this is right" operation.

6 years agoefl selection manager - fix x11 cnp from entry to plain text conversion
Carsten Haitzler (Rasterman) [Fri, 13 Apr 2018 14:25:52 +0000 (23:25 +0900)]
efl selection manager - fix x11 cnp from entry to plain text conversion

use sel data not input data/size... copy & paste fix problem...
fixes 28d06f9a5fa2b639579ac43b590df485e0cfd650
@fix

6 years agoelua: get rid of lightuserdata for state retrieval
Daniel Kolesa [Fri, 13 Apr 2018 13:09:21 +0000 (15:09 +0200)]
elua: get rid of lightuserdata for state retrieval

In order to get around the lightuserdata 48-bit problem, split
the state pointer into two and reconstruct it later as necessary.

6 years agobuild: disable epoll and inotify checking outside linux
Daniel Kolesa [Fri, 13 Apr 2018 13:07:53 +0000 (15:07 +0200)]
build: disable epoll and inotify checking outside linux

While compat shims exist for systems outside linux to provide
both, these require extra linkage and we're currently not set
up for that, besides we have native kqueue impls where relevant
so it's probably best to just disable.

6 years agotests: fix iterator leak in evas mesh tests
Mike Blumenkrantz [Fri, 13 Apr 2018 10:39:44 +0000 (12:39 +0200)]
tests: fix iterator leak in evas mesh tests

Summary: Depends on D5911

Reviewers: stefan_schmidt

Reviewed By: stefan_schmidt

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5912

6 years agotests: remove sleep() calls from eo tests
Mike Blumenkrantz [Fri, 13 Apr 2018 10:34:08 +0000 (12:34 +0200)]
tests: remove sleep() calls from eo tests

Summary: fix T6854

Reviewers: stefan_schmidt

Reviewed By: stefan_schmidt

Subscribers: cedric

Maniphest Tasks: T6854

Differential Revision: https://phab.enlightenment.org/D5910

6 years agotests: split elm_test_init into separate test suite (inside elm_suite)
Mike Blumenkrantz [Fri, 13 Apr 2018 10:26:08 +0000 (12:26 +0200)]
tests: split elm_test_init into separate test suite (inside elm_suite)

Summary:
check whether init+shutdown work explicitly before proceeding to other
tests
Depends on D5913

Reviewers: stefan_schmidt

Reviewed By: stefan_schmidt

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5914

6 years agoRevert "elm entry - fix wayland bug by not updating selection until after mse up"
Derek Foreman [Fri, 13 Apr 2018 11:41:34 +0000 (20:41 +0900)]
Revert "elm entry - fix wayland bug by not updating selection until after mse up"

Summary:
This reverts commit b0a7c4b086b8b15615a81b811eda9b215608e854.

This is adding another EFL bug to work around an EFL bug.

Reviewers: raster

Subscribers: devilhorns, stefan_schmidt, bu5hm4n, thiepha, cedric

Differential Revision: https://phab.enlightenment.org/D5933

6 years agoefl selection manager - fix dnd end handling to only handle on dnd
Carsten Haitzler (Rasterman) [Fri, 13 Apr 2018 11:39:07 +0000 (20:39 +0900)]
efl selection manager - fix dnd end handling to only handle on dnd

dnd end handling was being done even if no dnd started, like an
ungrab. this caused side effects like stoping selections in their
tracks.

@fix

6 years agoefl_ui_focus_manager_calc: do not redirect requested calls
Marcel Hollerbach [Thu, 12 Apr 2018 16:07:36 +0000 (18:07 +0200)]
efl_ui_focus_manager_calc: do not redirect requested calls

this is not documented and does not make sense.

6 years agoefl_ui_focus_util: add helper for getting the highest manager objekt
Marcel Hollerbach [Thu, 12 Apr 2018 16:04:18 +0000 (18:04 +0200)]
efl_ui_focus_util: add helper for getting the highest manager objekt

6 years agoelm_suite: add test for request_move
Marcel Hollerbach [Thu, 12 Apr 2018 15:44:48 +0000 (17:44 +0200)]
elm_suite: add test for request_move

6 years agoefl_ui_focus_manager_calc: a node from focus_stack is always normal node.
YeongJong Lee [Thu, 12 Apr 2018 12:35:55 +0000 (14:35 +0200)]
efl_ui_focus_manager_calc: a node from focus_stack is always normal node.

Summary:
focus_stack can't store logical node. we don't need to check a node from
focus_stack is not normal.
see line 1613

Reviewers: bu5hm4n

Reviewed By: bu5hm4n

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5927

6 years agoefl_ui_focus_manager_calc: remove needless check for normal node
YeongJong Lee [Thu, 12 Apr 2018 12:35:38 +0000 (14:35 +0200)]
efl_ui_focus_manager_calc: remove needless check for normal node

Summary:
a node in focus stack is always normal node. we don't need to check whether
node is normal.

see also, 97ec48434f59c18c5fb60a9f9eed394c82fb5f91

Reviewers: bu5hm4n

Reviewed By: bu5hm4n

Subscribers: cedric

Differential Revision: https://phab.enlightenment.org/D5930

6 years agoefl_ui_focus_manager: make request_move more powerfull
Marcel Hollerbach [Thu, 12 Apr 2018 12:28:31 +0000 (14:28 +0200)]
efl_ui_focus_manager: make request_move more powerfull

it turns out that it is useful and needed (for future patches) to
request moves for nodes that are not focused currently. It is also
needed to request a move that might end up in a logical node.

6 years agoefl_ui_focus_manager_calc: only call next when the node is unexpected
Marcel Hollerbach [Thu, 12 Apr 2018 12:27:22 +0000 (14:27 +0200)]
efl_ui_focus_manager_calc: only call next when the node is unexpected

ref T6793

6 years agomigration: turn off tizen only elementary
Wonki Kim [Mon, 28 May 2018 09:59:56 +0000 (18:59 +0900)]
migration: turn off tizen only elementary

Change-Id: Id507d157482d2446b2b01c4b7d9f5100c659c892
Signed-off-by: Wonki Kim <wonki_.kim@samsung.com>
6 years agoModify initial visibility of a extn_socket
Wonki Kim [Wed, 16 May 2018 01:42:37 +0000 (10:42 +0900)]
Modify initial visibility of a extn_socket

Summary:
extn socket should be rendered even if it is not visible to answer client's requests

extn_socket was rendered even if it's not visible for the first render.
however, it's not rendered as expected because there has been changes for ecore evas idle enterer.
so that this patch changes visibility of extn_socket to answer client's requests immediately.

ref D6054

Reviewers: zmike

Subscribers: cedric

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D6167

Change-Id: I723546ff4e7899f6056f9dd21e1f05662ae20a3d
Signed-off-by: Wonki Kim <wonki_.kim@samsung.com>
6 years agoecore_evas_extn: Remove aka latch logic on fn_prepare function
Wonki Kim [Tue, 15 May 2018 15:29:54 +0000 (00:29 +0900)]
ecore_evas_extn: Remove aka latch logic on fn_prepare function

Summary:
current extn_socket_prepare function logic could run into dead-end
because extnbuf_lock return null on failure of taking a lock.

this patch removes aka latch logic that guide the function to dead-end.

Reviewers: cedric, raster, zmike

Subscribers: zmike

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D6054

Change-Id: I55b7292455293f18e215a79387d6d605d8203565
Signed-off-by: Wonki Kim <wonki_.kim@samsung.com>
6 years agoecore_evas_extn: Modify not to send a message if there is nothing to draw
Wonki Kim [Wed, 9 May 2018 14:46:00 +0000 (23:46 +0900)]
ecore_evas_extn: Modify not to send a message if there is nothing to draw

Summary:
Seding ipc messages on ews_update_image(which is invoked as a callback of EVAS_RENDER_POST)
can makes ecore main loop awake, because sending ipc messages makes message_queue not empty.

so that this patch prevent seding ipc logic if there is nothing to draw.

Reviewers: cedric, raster

Reviewed By: raster

Subscribers: zmike

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D6055

Change-Id: Ibd11e40c5d1edaf2e94a8a00ae309d05c61c4ea1

6 years agoecore_event: fix backword compatibility bug for ecore event 47/178447/2
Hosang Kim [Tue, 8 May 2018 11:42:47 +0000 (20:42 +0900)]
ecore_event: fix backword compatibility bug for ecore event

Before ECORE_EVENT_NONE is 0 and it is not used event type.
https://phab.enlightenment.org/D5940

Change-Id: Ida409e865e52176ff523df707ed0ce1feec33e7d
(cherry picked from commit 8492b1c7801c2fba99c9cc90ecdd3320d00ec1f9)

6 years agoevas_object_box: fix compatibility bug for calculating box items 68/178568/1
Hosang Kim [Thu, 10 May 2018 11:51:29 +0000 (20:51 +0900)]
evas_object_box: fix compatibility bug for calculating box items

previous commit: 0f08ee686ecf91d7815f35ddc8481fd7d2936988
evas box - fix longstanding bug where min size is miscalculated

@tizen_fix

Change-Id: Id36d72e22d6da818f077662501ebbb78d3094962
Signed-off-by: Hosang Kim <hosang12.kim@samsung.com>
6 years agoefreet: disable unnecessary parts. 24/177324/3
Hosang Kim [Fri, 27 Apr 2018 06:04:51 +0000 (15:04 +0900)]
efreet: disable unnecessary parts.

Tizen doesn't use efreet mime and freedesktop icon.

@Tizen

Change-Id: I0872ccb453ca88978bd88f38d8a38290f13de2e2

6 years agogenlist: prevent memory leak in item class update 74/178174/1
SangHyeon Lee [Tue, 8 May 2018 11:37:23 +0000 (20:37 +0900)]
genlist: prevent memory leak in item class update

Change-Id: I22c823961f9f187c559c1816788b85348554bf28

6 years agogenlist : fix ABI break in gen common 59/178159/1
SangHyeon Lee [Tue, 8 May 2018 10:17:00 +0000 (19:17 +0900)]
genlist : fix ABI break in gen common

Change-Id: I469974e5885844e9da7cad5c33cfcf540904bedf
(cherry picked from commit c5183bc2e956e278d27c9e56c02385a576b543aa)

6 years agocursor: fix ABI break in widget item cursor APIs 47/178147/2
SangHyeon Lee [Thu, 3 May 2018 10:40:45 +0000 (19:40 +0900)]
cursor: fix ABI break in widget item cursor APIs

fix ABI break in widget item cursor APIs
by efl_ui_cursor interface implementing.

item's view were not elementary widget mostly,
reimplementing efl_ui_cursor on efl_ui_widget
cannot resolve item's cursor API.

Change-Id: Ie5cf051d5d22e10f4c11ef8747d9bd9a3fccae1a
Signed-off-by: SangHyeon Lee <sh10233.lee@samsung.com>
(cherry picked from commit 0190020675ab40f7db6882f42ce4ecf46e5b53c5)

6 years agogen : fix ABI break on Elm_Gen_Item_Class 46/178146/1
SangHyeon Lee [Thu, 3 May 2018 03:33:00 +0000 (12:33 +0900)]
gen : fix ABI break on Elm_Gen_Item_Class

boolean value homogeous was been placed between middle of
delete_me and item_style in Tizen_4.0
so we have to maintain same position about the values On
the public structure to prevent ABI break.

Change-Id: I50e5899c0c0e2e889f2c0de0452646b6c6c57464
(cherry picked from commit 404f34981f7ac75f63524adf838777abb2131089)

6 years agoelm_object : Support to translatable_part_text of legacy widget 39/177939/3
JunsuChoi [Fri, 4 May 2018 11:07:18 +0000 (20:07 +0900)]
elm_object : Support to translatable_part_text of legacy widget

Summary:
this commit supports aliasing of legacy widget
about translatable_part_text_set/get.

Test Plan:
elm_object_domain_translatable_part_text_set(btn, "default", NULL, "sample_text");
const char* text = elm_object_translatable_part_text_get(btn, "default");

Reviewers: cedric, herb, id213sin, woohyun

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D6056

Change-Id: I0295a74eea971ea5b0cb6de7fa63dd70afac374f

6 years agoecore_con,ecore_evas,evas_main: Fix the static analysis issues 53/177853/1
Myoungwoon Roy, Kim [Fri, 4 May 2018 04:32:03 +0000 (13:32 +0900)]
ecore_con,ecore_evas,evas_main: Fix the static analysis issues

ecore_evas_buffer: fix an warnning for the unchecked return value
ecore_con: fix an warning for the uninitialized scalar variable(UNINT)
evas_main: fix structurally dead code

Change-Id: I28eaa41b8f14518c07898401c9c056b700b8fb59

6 years agoecore_ipc: Restores ability for checking negative port 58/177458/7
Myoungwoon Roy, Kim [Mon, 30 Apr 2018 07:28:19 +0000 (16:28 +0900)]
ecore_ipc: Restores ability for checking negative port

Restores ability for checking negative port on local/remote ipc connection

Change-Id: I973aeaf4e8e4b775a088efb3cebaaa2000dc8afc

6 years agoevas_out: resolve possible memory leak
Jongmin Lee [Fri, 27 Apr 2018 00:24:18 +0000 (09:24 +0900)]
evas_out: resolve possible memory leak

Change-Id: Ia9664526a5b8126dcea302b382f45328d9cbd089

6 years agoefl_selection_manager: resolve possible memory leak
Jongmin Lee [Fri, 27 Apr 2018 00:51:41 +0000 (09:51 +0900)]
efl_selection_manager: resolve possible memory leak

Change-Id: I91a728c349659ec40b852ff1e6b67c7110da8a63

6 years agoedje_util: Fix object isolate issue.
Woochan Lee [Thu, 26 Apr 2018 06:30:56 +0000 (15:30 +0900)]
edje_util: Fix object isolate issue.

@tizen_fix
Change-Id: I608727277a098e0a0a824675cb2bbc64194f382e

6 years agoevas textblock: prevent invalid read from a free'd cursor
Youngbok Shin [Wed, 25 Apr 2018 07:26:05 +0000 (16:26 +0900)]
evas textblock: prevent invalid read from a free'd cursor

After calling destructor of an object, render_pre function can be called for that object.
In this case, we need to handle pointer carefully.

@tizen_fix

Change-Id: I6ab50e88402892568b53a25622be621328c67823

6 years agoelm: resolve possible integer overflow
Jongmin Lee [Fri, 20 Apr 2018 08:24:39 +0000 (17:24 +0900)]
elm: resolve possible integer overflow

Change-Id: I811cb7ef2bfda7016e1c41c06442aeffbd2be558

6 years agoelementary bg: fix legacy bg scale policy
Youngbok Shin [Tue, 24 Apr 2018 09:54:16 +0000 (18:54 +0900)]
elementary bg: fix legacy bg scale policy

It has to reverted when D5763 is merged.
@tizen_fix

Change-Id: Id140579bb09d02ddd67a3b9e513ffa962f2e4597

6 years agomodify code to handle window position
Doyoun Kang [Tue, 24 Apr 2018 07:00:36 +0000 (16:00 +0900)]
modify code to handle window position

Change-Id: I8284f3b31f7ec3b4b14c6f0a6dd77a71a1ae666c

6 years agocbhm_helper: fix the memory leak
Taehyub Kim [Mon, 23 Apr 2018 11:57:17 +0000 (20:57 +0900)]
cbhm_helper: fix the memory leak

Change-Id: Icee9bcaf1b443371e0ba1dfa0bd7633b8ae587b2

6 years agoelementary bg: allow NULL pointers in legacy usage
Youngbok Shin [Mon, 23 Apr 2018 11:20:17 +0000 (20:20 +0900)]
elementary bg: allow NULL pointers in legacy usage

Summary:
It should not cause crash with NULL parameters.
@fix

Reviewers: woohyun, raster, cedric

Differential Revision: https://phab.enlightenment.org/D5979

Change-Id: I8b7fb6bdbe7dd6f491a5edefe6529acb755a0591

6 years agoelementary: bg - keep file path and key string for legacy bg widget
Youngbok Shin [Wed, 7 Mar 2018 01:56:56 +0000 (17:56 -0800)]
elementary: bg - keep file path and key string for legacy bg widget

Summary:
If a file path and key string was passed to elm_bg, we could get
the same file path and key string. Even if it failed to load the image file.
And the file path also remained its original form.
ex) Setting file path "~/image.png" => Getting file path "~/image.png"
                                       (Not "/home/user_name/image.png")

@fix

Test Plan: Included in elementary test suite.

Reviewers: jpeg, cedric, raster

Reviewed By: cedric

Subscribers: woohyun

Differential Revision: https://phab.enlightenment.org/D5823

Change-Id: Iadc615664d27832b5e232a150f0cd82f197defd2
Signed-off-by: Cedric Bail <cedric@osg.samsung.com>
6 years agoRevert "popup: disable line wrap set to elm_label"
Wonki Kim [Fri, 27 Apr 2018 07:08:07 +0000 (16:08 +0900)]
Revert "popup: disable line wrap set to elm_label"

This reverts commit 22b439213a0edf841abb6b23eea4332a02927343.

6 years agoUi layout: fix latest attempt to fix markup retreival
Daniel Hirt [Thu, 1 Feb 2018 10:57:43 +0000 (12:57 +0200)]
Ui layout: fix latest attempt to fix markup retreival

Commits c07a40c745c5df1b1f6f0bbf666b233d8d072ca7 and
429e19563d15a02d3417ff8fffca994ce6685249 tried to fix markup retrieval by
creating a different code path for the setter and getter. However, legacy code
paths needs to be taken into consideration.
This commit fixes legacy widgets (like guide text that did not show any text
        in the entry at the top of 'elementary_test').

ref T6642

6 years agoevas_main: Add null exception handling 92/176992/5
Myoungwoon Roy, Kim [Tue, 24 Apr 2018 10:45:56 +0000 (19:45 +0900)]
evas_main: Add null exception handling

Need to check whether Evas_Public_Data data is null or not before dereferencing it.

Change-Id: I57aeab8114cf41591a47cf70a388ed634680d2d5

6 years agowidget: stop creating unused class. 47/177247/1
Hosang Kim [Thu, 26 Apr 2018 11:00:38 +0000 (20:00 +0900)]
widget: stop creating unused class.

The combobox class is unnecessary class.

Change-Id: I848e3aae2e029076cec252e8b2de7392a196d3fb

6 years agoscrl_iface: free adjust job when interface is freed 82/177182/1
SangHyeon Lee [Thu, 26 Apr 2018 06:30:15 +0000 (15:30 +0900)]
scrl_iface: free adjust job when interface is freed

Change-Id: I1613ef191dbf505517d8278758bc1aabe1be0b48
Signed-off-by: SangHyeon Lee <sh10233.lee@samsung.com>
6 years agotizen audio: cherry-pick missing code from tizen branch 15/177115/4
Wonki Kim [Wed, 25 Apr 2018 11:05:57 +0000 (20:05 +0900)]
tizen audio: cherry-pick missing code from tizen branch

Change-Id: Id1fae479b6b0e36bb49d9a32c72ebe5842150fed

6 years agoFix a error cased by rebasing a patch f1ac23d
Wonki Kim [Wed, 25 Apr 2018 04:24:12 +0000 (13:24 +0900)]
Fix a error cased by rebasing a patch f1ac23d

sd->subs is needed to be replaced by null while freeing it.

Change-Id: Ie45200f0858c7e2280557955452bcb53d26e2543
Signed-off-by: Wonki Kim <wonki_.kim@samsung.com>
6 years agoRevert "ecore_wl2: Fix ecore_wl2_window_output_find"
Wonki Kim [Tue, 24 Apr 2018 13:19:18 +0000 (22:19 +0900)]
Revert "ecore_wl2: Fix ecore_wl2_window_output_find"

This reverts commit 2f474a3e7d13b99fdfa6b058ba1bbd2b73a8732f.

6 years agoefl_ui_win: Modified to use efl_config->win_no_border in _elm_win_finalize_internal
Jeonghyun Yun [Mon, 16 Apr 2018 08:14:51 +0000 (17:14 +0900)]
efl_ui_win: Modified to use efl_config->win_no_border in _elm_win_finalize_internal

@tizen_fix

Change-Id: Ia42a1fefc7fbfe4cebc91398c66c4b2c73fb61d0
Signed-off-by: Jeonghyun Yun <jh0506.yun@samsung.com>
6 years agoRevert "efl_ui_win: move setting borderless code."
Wonki Kim [Tue, 24 Apr 2018 06:32:16 +0000 (15:32 +0900)]
Revert "efl_ui_win: move setting borderless code."

This reverts commit 36a493b7a50908f1abe84bdfca272d7cbdf51e54.

6 years agoFix a bug about the value of map point colors 07/176907/1
Myoungwoon Roy, Kim [Tue, 24 Apr 2018 06:47:30 +0000 (15:47 +0900)]
Fix a bug about the value of map point colors

evas_map_point_color_get returns wrong RGBA values always because of error label.

Change-Id: I1583d4a01039d69775729ac5cc5930a3be5e69e1

6 years agoatspi : fix memory leak
JunsuChoi [Mon, 23 Apr 2018 10:17:07 +0000 (19:17 +0900)]
atspi : fix memory leak

   If return at check iter_entry and iter_dict,
   it occured memory leak form ret

Change-Id: Ia0cb9aeeab8bdd792257d8257c27d56de1c62ec6

6 years agoRevert "ecore_wl2: ignore first configure cb after client change the window size...
Jiyoun Park [Mon, 23 Apr 2018 06:51:24 +0000 (15:51 +0900)]
Revert "ecore_wl2: ignore first configure cb after client change the window size explictly"

This reverts commit 4c66dd5f9239f8c5324ffeed93f4ffc89af3a010.

Change-Id: Ic62469932395fd36ac3df5741aeee2c677a1702e
Signed-off-by: Jiyoun Park <jy0703.park@samsung.com>
6 years agoelementary: bg - keep file path and key string for legacy bg widget
Youngbok Shin [Wed, 7 Mar 2018 01:56:56 +0000 (17:56 -0800)]
elementary: bg - keep file path and key string for legacy bg widget

Summary:
If a file path and key string was passed to elm_bg, we could get
the same file path and key string. Even if it failed to load the image file.
And the file path also remained its original form.
ex) Setting file path "~/image.png" => Getting file path "~/image.png"
                                       (Not "/home/user_name/image.png")

@fix

Test Plan: Included in elementary test suite.

Reviewers: jpeg, cedric, raster

Reviewed By: cedric

Subscribers: woohyun

Differential Revision: https://phab.enlightenment.org/D5823

Change-Id: Iadc615664d27832b5e232a150f0cd82f197defd2
Signed-off-by: Cedric Bail <cedric@osg.samsung.com>
6 years agoelementary layout: just return false when the given pos is negative
Youngbok Shin [Sat, 21 Apr 2018 09:15:27 +0000 (18:15 +0900)]
elementary layout: just return false when the given pos is negative

The function for this feature in Edje starts supporting negative pos.
The negative pos will be used for inserting the given object at backward position.

But, in Tizen 4.0, the document for this function described pos has to be positive or zero.
Actually, it failed when pos was negative.
So, for backward compatibility, it must return FALSE in this case.

@tizen_fix

Change-Id: I004af7839976a94b5e4e68a64c795689f0a25ca3

6 years agoelementary: add additional class names for legacy API calls
Youngbok Shin [Mon, 23 Apr 2018 04:55:18 +0000 (13:55 +0900)]
elementary: add additional class names for legacy API calls

It is a temporary patch before updating code according to the latest upstream code.

@tizen_fix

Change-Id: Ie44327234d2d52418251a5b1603e74e837c1d2bb

6 years agoecore_wl2: ignore first configure cb after client change the window size explictly
Jiyoun Park [Sat, 21 Apr 2018 18:18:20 +0000 (03:18 +0900)]
ecore_wl2: ignore first configure cb after client change the window size explictly

aux_hint: wm.policy.win.user.geometry 1
if client used this hint, client can resize window size even thought basic window.
But client can get 1X1 configure size after set this hint.

we need to analyze the whole size negotiation process between client and server.

problem case:
ecore_evas_new: 1X1 (client->server)
ecore_evas_resize: 1920X1080 (client->server)
configure 1X1 (server->client)
client change the size 1X1 even thought they already request 1920X1080

this is hot fix because it only can ignore first configure after client size set.
we maybe need to use the serial value of ecore_wl2_win.

Change-Id: I1961a41a31e94e76f89d8db9fdb7793bd83d4ba1

6 years agoprogressbar: progress pulse policy is diffrent with spin.
Hosang Kim [Sat, 21 Apr 2018 09:39:07 +0000 (18:39 +0900)]
progressbar: progress pulse policy is diffrent with spin.

Change-Id: I4049a67a91e018df1cbe320420b98ed579e8dd7e

6 years agoelementary button: Do requested job even if a part name is improper for that job.
Youngbok Shin [Sat, 21 Apr 2018 05:57:47 +0000 (14:57 +0900)]
elementary button: Do requested job even if a part name is improper for that job.

When a text set is called for "elm.swallow.content", it will fail because of its part name.
It has to allow text set calls even the given part name is improper.

@tizen_fix

Change-Id: Ice95503363e7a5b1eb02fd1abae0996a7c1b8973