platform/upstream/efl.git
4 years agoecore-wl2: Move ecore_wl2_buffer_wl_buffer_get function to be internal
Christopher Michael [Tue, 5 Nov 2019 14:12:21 +0000 (09:12 -0500)]
ecore-wl2: Move ecore_wl2_buffer_wl_buffer_get function to be internal

As this function is only used in 1 place and will likely never be used
outside of the modular dmabuf engine, we can move it to be internal
only.

ref T8013

4 years agoecore-wl2: Move ecore_wl2_buffer_destroy function to be internal
Christopher Michael [Tue, 5 Nov 2019 14:11:56 +0000 (09:11 -0500)]
ecore-wl2: Move ecore_wl2_buffer_destroy function to be internal

As this function is only used in 1 place and will likely never be used
outside of the modular dmabuf engine, we can move it to be internal
only.

ref T8013

4 years agoecore-wl2: Move ecore_wl2_buffer_create function to be internal
Christopher Michael [Tue, 5 Nov 2019 14:11:35 +0000 (09:11 -0500)]
ecore-wl2: Move ecore_wl2_buffer_create function to be internal

As this function is only used in 1 place and will likely never be used
outside of the modular dmabuf engine, we can move it to be internal
only.

ref T8013

4 years agoecore-wl2: Move ecore_wl2_buffer_init function to be internal
Christopher Michael [Tue, 5 Nov 2019 14:10:48 +0000 (09:10 -0500)]
ecore-wl2: Move ecore_wl2_buffer_init function to be internal

As this function is only used in 1 place and will likely never be used
outside of the modular dmabuf engine, we can move it to be internal
only.

ref T8013

4 years agoecore-wl2: Move ecore_wl2_window_damage function to be internal
Christopher Michael [Tue, 5 Nov 2019 14:09:14 +0000 (09:09 -0500)]
ecore-wl2: Move ecore_wl2_window_damage function to be internal

As this function is only used in 1 place and will likely never be used
outside of the modular dmabuf engine, we can move it to be internal
only.

ref T8013

4 years agoecore-wl2-dmabuf: Add include for internal functions
Christopher Michael [Tue, 5 Nov 2019 14:08:23 +0000 (09:08 -0500)]
ecore-wl2-dmabuf: Add include for internal functions

4 years agoecore-wl2: Move variables to be above functions
Christopher Michael [Tue, 5 Nov 2019 13:53:29 +0000 (08:53 -0500)]
ecore-wl2: Move variables to be above functions

NB: No functional changes

4 years agodocs: Efl.Ui.Table_Static
Xavi Artigas [Tue, 5 Nov 2019 10:31:29 +0000 (11:31 +0100)]
docs: Efl.Ui.Table_Static

The need for this class is still being discussed in
https://phab.enlightenment.org/T8206
but at least it is a bit more clear what its purpose is.

4 years agoevas filter: Implement inverse color filter
Shinwoo Kim [Tue, 5 Nov 2019 01:44:58 +0000 (10:44 +0900)]
evas filter: Implement inverse color filter

Summary: This is the first version of inverse color filter both GL and SW.

Test Plan:
1. Create filter_example with following .
efl_gfx_filter_program_set(image, "inverse_color ()", "inverse color");

2. Run.
./filter_example (Use ELM_ACCEL=gl for GL engine)

Reviewers: Hermet, jsuya

Reviewed By: Hermet

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10335

4 years agobuild: add ecore_input_device
Hosang Kim [Thu, 7 Nov 2019 08:05:21 +0000 (17:05 +0900)]
build: add ecore_input_device

@tizen_only
Change-Id: I961f38268cfdae293342efc2c32b345285c6320e

4 years agobuild: fix build error
Hosang Kim [Thu, 7 Nov 2019 07:09:42 +0000 (16:09 +0900)]
build: fix build error

Change-Id: Ie4581daaeb1ceb68fc918b9b744758fc3717923f

4 years agocsharp: Changing visibility of nested class/struct
Bruno da Silva Belo [Mon, 4 Nov 2019 17:40:18 +0000 (14:40 -0300)]
csharp: Changing visibility of nested class/struct

Reviewers: felipealmeida, lauromoura, YOhoho

Reviewed By: YOhoho

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8393

Differential Revision: https://phab.enlightenment.org/D10588

4 years agodocs: Polish Efl.Canvas.Animation_Group and sons
Xavi Artigas [Wed, 30 Oct 2019 14:26:08 +0000 (15:26 +0100)]
docs: Polish Efl.Canvas.Animation_Group and sons

Summary: These docs were almost empty.

Reviewers: bu5hm4n, zmike, cedric, Jaehyun_Cho

Reviewed By: cedric

Subscribers: #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10564

4 years agoefl_canvas_animation: improve generally
Marcel Hollerbach [Mon, 4 Nov 2019 12:44:29 +0000 (13:44 +0100)]
efl_canvas_animation: improve generally

Summary:
to be taken over by doccop
Depends on D10559

Reviewers: Jaehyun_Cho, bu5hm4n

Reviewed By: bu5hm4n

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10560

4 years agodocs: Polish Efl.Canvas.Animation_*
Xavi Artigas [Mon, 4 Nov 2019 12:06:41 +0000 (13:06 +0100)]
docs: Polish Efl.Canvas.Animation_*

Summary:
These docs were lacking a lot of detail.

Depends on D10562

Reviewers: bu5hm4n, cedric, zmike, Jaehyun_Cho

Reviewed By: cedric

Subscribers: #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10565

4 years agoefl_canvas_animation_translate: move from x&y to container types
Marcel Hollerbach [Wed, 30 Oct 2019 14:30:20 +0000 (15:30 +0100)]
efl_canvas_animation_translate: move from x&y to container types

Summary:
this now uses Eina.Position2D so its easier to pass arround other
positions.

ref T8288
Depends on D10559

Reviewers: segfaultxavi

Reviewed By: segfaultxavi

Subscribers: segfaultxavi, cedric, #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8288

Differential Revision: https://phab.enlightenment.org/D10562

4 years agoefl_canvas_animation_scale: move from 2 doubles to vectors
Marcel Hollerbach [Wed, 30 Oct 2019 14:30:18 +0000 (15:30 +0100)]
efl_canvas_animation_scale: move from 2 doubles to vectors

Summary:
with this the passing of positions is getting more easy. Additionally,
Reading the API call gets easier, as parameters are semantically
grouped.
Depends on D10558

Reviewers: segfaultxavi, Jaehyun_Cho

Reviewed By: segfaultxavi

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8288

Differential Revision: https://phab.enlightenment.org/D10559

4 years agoefl_canvas_animation: be more explicit with errors
Marcel Hollerbach [Wed, 30 Oct 2019 14:30:09 +0000 (15:30 +0100)]
efl_canvas_animation: be more explicit with errors

Summary:
with this commit invalid values are not accepted silently anymore. But
rather a error will be raised.
Depends on D10350

Reviewers: segfaultxavi, Jaehyun_Cho

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8288

Differential Revision: https://phab.enlightenment.org/D10558

4 years agoefl_canvas_animation_rotate: adjust API
Marcel Hollerbach [Wed, 30 Oct 2019 14:29:59 +0000 (15:29 +0100)]
efl_canvas_animation_rotate: adjust API

Summary:
in task T8288 we concluded that a few APIs need to be adjusted in order
to stabelize animation classes at some point. This also adds a new macro
to eina in order to create EINA_VECTOR2 values more easily.

ref T8288

Reviewers: Jaehyun_Cho, segfaultxavi, zmike

Reviewed By: zmike

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8288

Differential Revision: https://phab.enlightenment.org/D10350

4 years agoeo_mono: make Efl.EventDescription, Efl.Event, Efl.EventCb internal
Yeongjong Lee [Fri, 1 Nov 2019 20:13:57 +0000 (17:13 -0300)]
eo_mono: make Efl.EventDescription, Efl.Event, Efl.EventCb internal

Summary:
Hide struct and delegate related to `IntPtr`.

Depends on D10585
Depends on D10586

Test Plan: meson setup -Dbindings=mono,cxx -Dmono-beta=true

Reviewers: lauromoura, Jaehyun_Cho

Reviewed By: lauromoura

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10587

4 years agomono: blacklist efl_thread
Yeongjong Lee [Fri, 1 Nov 2019 20:03:56 +0000 (17:03 -0300)]
mono: blacklist efl_thread

Summary:
C# developers are already familar with C# Thread`System.Threading.Thread`, We
don't need to provide Wrapped `Efl.Thread` class.
Also, we can't ensure compatibility between C# Thread and EFL Thread.

Test Plan: meson setup -Dbindings=mono,cxx -Dmono-beta=true

Reviewers: lauromoura, Jaehyun_Cho

Reviewed By: lauromoura

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10586

4 years agomono: blacklist functions related to native event
Yeongjong Lee [Fri, 1 Nov 2019 19:01:49 +0000 (16:01 -0300)]
mono: blacklist functions related to native event

Summary:
`efl_event_callback_forwarder_priority_del`
=> It can be replaced with `obj.XXXEvent -= callback;`.
Furthermore, `efl_event_callback_forwarder_priority_add` is already in blacklist.

`efl_ui_widget_input_event_handler`
=> It can be replaced with `obj.DownEvent`, `obj.UpEvent` and `obj.PointerWhellEvent`.

`efl_access_object_event_handler_add`
`efl_access_object_event_handler_del`
`efl_access_object_event_emit`
=> They are functions to handle global event related to access(E.g. `elm_atspi_bridge`).
It should be generated to `static event` in C#.

Test Plan: meson setup -Dbindings=mono,cxx -Dmono-beta=true

Reviewers: lauromoura, Jaehyun_Cho

Reviewed By: lauromoura

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10585

4 years agocsharp: Refactor MarshalOwn
Lauro Moura [Fri, 1 Nov 2019 13:04:04 +0000 (10:04 -0300)]
csharp: Refactor MarshalOwn

Summary:
Instead of using the empty interface as tag, split MarshalOwn into two
marshalers that can be used directly.

Fixes T8395 (CA1040)

Test Plan: no new functionality. Use existing tests

Reviewers: felipealmeida, brunobelo, segfaultxavi, YOhoho

Reviewed By: YOhoho

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8395

Differential Revision: https://phab.enlightenment.org/D10457

4 years agoRevert "elm/genlist: don't process entire item queue on each item add"
Carsten Haitzler (Rasterman) [Fri, 1 Nov 2019 10:12:30 +0000 (10:12 +0000)]
Revert "elm/genlist: don't process entire item queue on each item add"

First - the big problem. This breaks enlightenment's bluez5 popup. it
does a sortyed inert using the item data and the item data for one of
the itmes to compare in _cb_insert_cmp() in e_mod_popup.c when it
calls elm_object_item_data_get(0 returns a junk ptr for the item data
after this patch. i haven't managed to figure out exactly why in my
last 30 mins of looking.

But a closer look... this disables immediate processing of:

1. the first block of items (32items) which was intended so for
short/small lists you have some content by the time you go to the
first frame, and at least the first block of itso you seem to have
visual contnt and not a blank list until idlers can process further
content. so the patch being reverted would have gotten rid of this
logic that gets you content as opposed to blank:

  while ((sd->queue) && ((!sd->blocks) || (!sd->blocks->next)))

2. if it's a homogenous list, all items have the same size so we do
have to realize the first item of each class type but ONLY that one.
further items should not need realizing as we can ASSUME the height to
be the same as the first item... that's the point of homogenous +
compress lists - all items of the same class have the same height and
width thus shortcutting further need to calculate nd realize. if we
are reizing everything in a homogenous list then the issue lies with
something going wrong with this above logic. we shokuld be able to
handle such lists super fastif this logic was working.
that's the 2nd while:

   while ((sd->queue) && (sd->blocks) &&
             (sd->homogeneous) && (sd->mode == ELM_LIST_COMPRESS))

so overall, this should not have been realizing every item. either
just the first block of 32, OR just the first item of any class and
thus assume all further items are the same size without realizing
being needed. if these broke then the solution is not commenting this
out but finding out why this logic is breaking :)

and not to mention... this commenting out also caused segfaults in
existing applications which are doing the right thing. perhaps the
sorting logic also needed updating as well if this above is commented out...
but i didn't have time to chase it more than this.

---

This reverts commit 0777b74f07857c86408bc0929e3391ced0e098e4.

4 years agoecore_event: Remove unused goto define
JunsuChoi [Fri, 1 Nov 2019 03:46:24 +0000 (12:46 +0900)]
ecore_event: Remove unused goto define

Summary:
   For remove -Wunused-label warning

Test Plan: N/A

Reviewers: Hermet, kimcinoo, YOhoho

Reviewed By: YOhoho

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10584

4 years agomono: blacklist elm_general.eot
Yeongjong Lee [Thu, 31 Oct 2019 21:30:23 +0000 (18:30 -0300)]
mono: blacklist elm_general.eot

Summary:
Prevent generating legacy eot.cs
Depends on D10580

Test Plan: meson setup -Dbindings=mono,cxx -Dmono-beta=true

Reviewers: lauromoura, Jaehyun_Cho

Reviewed By: lauromoura

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10581

4 years agomono: blacklist efl_ui_text_context_menu_item_add
Yeongjong Lee [Thu, 31 Oct 2019 21:25:32 +0000 (18:25 -0300)]
mono: blacklist efl_ui_text_context_menu_item_add

Summary: Blacklist a method which has `IntPtr` parameters.

Reviewers: woohyun, segfaultxavi, YOhoho, felipealmeida

Reviewed By: felipealmeida

Subscribers: segfaultxavi, zmike, q66, bu5hm4n, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10579

4 years agomono: remove duplicated EinaFreeCb delegate
Yeongjong Lee [Thu, 31 Oct 2019 21:21:42 +0000 (18:21 -0300)]
mono: remove duplicated EinaFreeCb delegate

Summary: Use Eina.Claabacks.EinaFreeCb instead.

Reviewers: lauromoura

Reviewed By: lauromoura

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10577

4 years agoefl: fix include due to stale files.
Cedric BAIL [Thu, 31 Oct 2019 19:43:16 +0000 (15:43 -0400)]
efl: fix include due to stale files.

Reviewers: zmike

Reviewed By: zmike

Subscribers: #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10583

4 years agoefl_ui_text: remove elm_general.eot
Yeongjong Lee [Thu, 31 Oct 2019 17:38:08 +0000 (13:38 -0400)]
efl_ui_text: remove elm_general.eot

Summary: remove legacy dependency in eo file.

Test Plan: ninja

Reviewers: woohyun, Jaehyun_Cho, zmike

Reviewed By: zmike

Subscribers: zmike, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10580

4 years agoevas: move watching over destruction of device to custom logic due to high use.
Cedric Bail [Thu, 31 Oct 2019 17:20:52 +0000 (13:20 -0400)]
evas: move watching over destruction of device to custom logic due to high use.

Summary:
This specific EFL_EVENT_DEL handler was registered thousand of time alone on
an Evas device. Potential solution are to improve handling of this kind of large
scale callback logic or just not take that path. I find it easier to have a custom
code in this case to handle the destruction of Evas device and avoid this
thousand of callback.

Depends on D10492

Reviewers: zmike, raster, bu5hm4n, Hermet

Reviewed By: zmike

Subscribers: #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8321

Differential Revision: https://phab.enlightenment.org/D10493

4 years agoevas: deprecate evas_device_parent_set.
Cedric Bail [Thu, 31 Oct 2019 17:20:45 +0000 (13:20 -0400)]
evas: deprecate evas_device_parent_set.

Summary:
It is unlikely that the code was working before and it was a bad idea
anyway. There is no user of this API in EFL.

Depends on D10490

Reviewers: zmike, raster, bu5hm4n, Hermet

Reviewed By: zmike

Subscribers: #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8321

Differential Revision: https://phab.enlightenment.org/D10491

4 years agoevas: move exposed internal structure of Efl_Input_Device to be private.
Cedric Bail [Thu, 31 Oct 2019 17:20:39 +0000 (13:20 -0400)]
evas: move exposed internal structure of Efl_Input_Device to be private.

Summary: Depends on D10489

Reviewers: zmike, raster, bu5hm4n, Hermet

Reviewed By: zmike

Subscribers: #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8321

Differential Revision: https://phab.enlightenment.org/D10490

4 years agoevas: move efl_input_device into evas/Efl_Canvas.h
Cedric Bail [Thu, 31 Oct 2019 17:20:33 +0000 (13:20 -0400)]
evas: move efl_input_device into evas/Efl_Canvas.h

Summary:
The internal and the API we would like is mostly a canvas API. A lot of the code
in evas is working around the fact that efl_input_device is not defined inside Evas.
This patch is the first step to try to clean this up.

Depends on D10487

Reviewers: zmike, raster, bu5hm4n, Hermet

Reviewed By: zmike

Subscribers: #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8321

Differential Revision: https://phab.enlightenment.org/D10488

4 years agotests/eina: use #ifdef guard for block which can only occur inside ifdef
Mike Blumenkrantz [Wed, 9 Oct 2019 17:20:43 +0000 (13:20 -0400)]
tests/eina: use #ifdef guard for block which can only occur inside ifdef

this value is set in the corresponding #ifdef block above, so use the same
guard here to make code more readable

CID 1400948

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10321

4 years agoeo_base_class: move shift init to the first call
Marcel Hollerbach [Thu, 31 Oct 2019 16:48:27 +0000 (12:48 -0400)]
eo_base_class: move shift init to the first call

Summary:
this cannot be evalulated in compile time, so this must be evalulated in
runtime, at the first call.

This should fix OSX build.

Co-authored-by: Cedric Bail <cedric.bail@free.fr>
Reviewers: zmike, cedric, raster

Reviewed By: zmike

Subscribers: #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10582

4 years agoeo: add debug ability to detect long chain of event handler.
Cedric Bail [Fri, 18 Oct 2019 16:24:47 +0000 (09:24 -0700)]
eo: add debug ability to detect long chain of event handler.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D10484

4 years agoeo: reduce memory use to track event registration and callback call.
Cedric Bail [Thu, 17 Oct 2019 19:09:14 +0000 (12:09 -0700)]
eo: reduce memory use to track event registration and callback call.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D10482

4 years agoecore: remove custom code for generating Eina_Future_Schedule attached on Efl.Loop.
Cedric Bail [Wed, 16 Oct 2019 23:47:47 +0000 (16:47 -0700)]
ecore: remove custom code for generating Eina_Future_Schedule attached on Efl.Loop.

This leverage the new infrastructure from Eo that provide a scheduler for any event
attached to any object.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D10481

4 years agoeo: add infrastructure to attach an Eina_Future_Scheduler to any source of event.
Cedric Bail [Wed, 16 Oct 2019 23:47:10 +0000 (16:47 -0700)]
eo: add infrastructure to attach an Eina_Future_Scheduler to any source of event.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D10480

4 years agoeina: only the type need to be NULL to assume EINA_VALUE_EMPTY.
Cedric Bail [Wed, 16 Oct 2019 23:45:50 +0000 (16:45 -0700)]
eina: only the type need to be NULL to assume EINA_VALUE_EMPTY.

This avoid comparison with potentially uninitialized byte.

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D10479

4 years agoecore: remove unecessary code for Eina_Future scheduler.
Cedric Bail [Wed, 16 Oct 2019 19:14:20 +0000 (12:14 -0700)]
ecore: remove unecessary code for Eina_Future scheduler.

As we do not rely on legacy Ecore Event directly anymore, we do not
need to mind the shutting down of EFL.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D10477

4 years agoecore: remove unecessary field in data structure.
Cedric Bail [Wed, 16 Oct 2019 17:04:29 +0000 (10:04 -0700)]
ecore: remove unecessary field in data structure.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D10476

4 years agoeo: prevent unecessary callback call walk.
Cedric Bail [Wed, 16 Oct 2019 00:04:32 +0000 (17:04 -0700)]
eo: prevent unecessary callback call walk.

This patch introduce a small hash (64 or 32bits) that cache all existing
Efl_Event_Description callback handler registered on an object. It slightly
reduce the time needed to do an unecessary call and cost just a few bytes
per object.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D10475

4 years agoelementary: handle case when XFIXES is not available.
Cedric Bail [Thu, 31 Oct 2019 13:17:53 +0000 (09:17 -0400)]
elementary: handle case when XFIXES is not available.

Summary:
ECORE_X_EVENT_FIXES_SELECTION_NOTIFY is only initialized when XFIXES
is available. If ecore_event_handler_add is called with type == 0, it
will trigger an abort and elementary would not initialize properly.

Depends on D10491

Reviewers: zmike, raster, bu5hm4n, Hermet

Reviewed By: zmike

Subscribers: #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8321

Differential Revision: https://phab.enlightenment.org/D10492

4 years agoedje: load edje seat callback only when necessary.
Cedric Bail [Thu, 31 Oct 2019 13:17:46 +0000 (09:17 -0400)]
edje: load edje seat callback only when necessary.

Summary:
This reduce in elementary_test the number of callback registered on the
canvas from hundreds to around 10.

Depends on D10486

Reviewers: zmike, raster, bu5hm4n, Hermet

Reviewed By: zmike

Subscribers: #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8321

Differential Revision: https://phab.enlightenment.org/D10487

4 years agoedje: improve callback count on Evas canvas.
Cedric Bail [Thu, 31 Oct 2019 13:17:40 +0000 (09:17 -0400)]
edje: improve callback count on Evas canvas.

Summary:
This reduce by 3 the amount of callback registered on the canvas. Another
potential improvement would be to only register those callback if someone
is listening for a 'seat,*' event or if the edje file define seat filters.

Depends on D10484

Reviewers: zmike, raster, bu5hm4n, Hermet

Reviewed By: zmike

Subscribers: #reviewers, #committers

Tags: #efl

Maniphest Tasks: T8321

Differential Revision: https://phab.enlightenment.org/D10486

4 years agotests/spec: pre-create a ton of widgets during test init
Mike Blumenkrantz [Thu, 31 Oct 2019 13:02:56 +0000 (09:02 -0400)]
tests/spec: pre-create a ton of widgets during test init

Summary:
similar to existing mechanics for elm/efl_ui tests, this pre-creates
all the widgets used as test contents for all the test cases so they
can be forked and reused without needing to spend time in every single
test creating the same test contents
Depends on D10573

Reviewers: cedric

Reviewed By: cedric

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10574

4 years agotests/elm: add utils for detecting buffer engine and setting fork init cb
Mike Blumenkrantz [Thu, 31 Oct 2019 13:02:54 +0000 (09:02 -0400)]
tests/elm: add utils for detecting buffer engine and setting fork init cb

Summary:
this allows use of the global win object prior to actually being inside a
test for the purpose of doing additional setup
Depends on D10572

Reviewers: cedric

Reviewed By: cedric

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10573

4 years agotests/spec: use wrefs to automatically unset global pointers
Mike Blumenkrantz [Thu, 31 Oct 2019 13:02:52 +0000 (09:02 -0400)]
tests/spec: use wrefs to automatically unset global pointers

Summary:
these may end up being reused in the future, so unset them only on
object destruction
Depends on D10571

Reviewers: cedric

Reviewed By: cedric

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10572

4 years agotests/spec: assert success of content_set during setup
Mike Blumenkrantz [Thu, 31 Oct 2019 13:02:51 +0000 (09:02 -0400)]
tests/spec: assert success of content_set during setup

Summary: Depends on D10570

Reviewers: cedric

Reviewed By: cedric

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10571

4 years agotests/spec: store test class and reuse globally to avoid needing efl_isa
Mike Blumenkrantz [Thu, 31 Oct 2019 13:02:49 +0000 (09:02 -0400)]
tests/spec: store test class and reuse globally to avoid needing efl_isa

Summary: Depends on D10569

Reviewers: cedric

Reviewed By: cedric

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10570

4 years agotests/elm: add util function for detecting whether current test is forked
Mike Blumenkrantz [Thu, 31 Oct 2019 13:02:47 +0000 (09:02 -0400)]
tests/elm: add util function for detecting whether current test is forked

Summary:
this is useful for determining whether we can use extra hacks because nobody's
supervising our pointer (ab)use
Depends on D10568

Reviewers: cedric

Reviewed By: cedric

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10569

4 years agotests/spec: reduce looping of pack content tests
Mike Blumenkrantz [Thu, 31 Oct 2019 13:02:45 +0000 (09:02 -0400)]
tests/spec: reduce looping of pack content tests

Summary:
this was hugely over-testing before; simply verifying one item outside
the limit is sufficient here
Depends on D10567

Reviewers: cedric

Reviewed By: cedric

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10568

4 years agotests/spec: merge some loops in linear pack tests
Mike Blumenkrantz [Thu, 31 Oct 2019 13:02:44 +0000 (09:02 -0400)]
tests/spec: merge some loops in linear pack tests

Summary:
these are small loops, but iterating small loops hundreds of times leads to
big slowdowns

Reviewers: cedric

Reviewed By: cedric

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10567

4 years agoevas - revert evas variation sequence support - out of bound accesses
Carsten Haitzler (Rasterman) [Thu, 31 Oct 2019 12:21:59 +0000 (12:21 +0000)]
evas - revert evas variation sequence support - out of bound accesses

This code is filled with out of bounds accesses now after the reverted
patch. All those base_char+1, itr+1 etc. in
evas_common_font_query_run_font_end_get() are accessing BEYOND
the end of the run. textgrid shows this instantly to fall over as it
uses single unicode codepoint chars with no nul terminator. As this
api takes an explicit run_len we should never access beyond the end of
the run_len.

Please revisit this code and keep in mind proper memory/bounds
accessing. If there was ano run_len and it assumed strings were
regular strings that had to be nul terminated... then it might be ok,
but not here.

of course if i put in guards for these +1's then it ends up in
infintie loops, so enough debugging and send it back for a rethink. :)

....

Revert "evas_object_textblock: add support for variation sequences"

This reverts commit 46f2d8acdcda3f374c9e393ecb734ff9d00fef7d.

4 years agocsharp: space after keywords.
Bruno da Silva Belo [Thu, 31 Oct 2019 10:45:01 +0000 (07:45 -0300)]
csharp: space after keywords.

Reviewers: felipealmeida, lauromoura, YOhoho

Reviewed By: lauromoura

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10576

4 years agotests/efl_app: add port of intensive timer recursion
Mike Blumenkrantz [Tue, 29 Oct 2019 17:08:21 +0000 (13:08 -0400)]
tests/efl_app: add port of intensive timer recursion

this is roughly the same as the similarly-named ecore_timer unit test

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10556

4 years agotests/evas: prune directory scan in mesh loader/saver test
Mike Blumenkrantz [Tue, 29 Oct 2019 14:07:53 +0000 (10:07 -0400)]
tests/evas: prune directory scan in mesh loader/saver test

this test should use a hardcoded list of files, but until it does we
can at least attempt to reduce failure rates which may occur from random
garbage in the tree

ref T6857

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10552

4 years agotests/evas: replace macro with function in mesh loader/saver test
Mike Blumenkrantz [Tue, 29 Oct 2019 14:05:06 +0000 (10:05 -0400)]
tests/evas: replace macro with function in mesh loader/saver test

this was totally impossible to debug

no functional changes

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10551

4 years agotests/ecore: add intensive test for recursive timer usage
Mike Blumenkrantz [Mon, 28 Oct 2019 19:50:02 +0000 (15:50 -0400)]
tests/ecore: add intensive test for recursive timer usage

this recurses the mainloop to a depth of 8, continually creating and
triggering timers as it progresses and tracking the states to ensure that
everything is working as expected regardless of depth

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10547

4 years agotests/ecore: do not create global test timeout timer if running unforked
Mike Blumenkrantz [Mon, 28 Oct 2019 19:03:28 +0000 (15:03 -0400)]
tests/ecore: do not create global test timeout timer if running unforked

this is a pita for debugging

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10546

4 years agoecore/timer: correctly handle recursive deletion of legacy timers
Mike Blumenkrantz [Mon, 28 Oct 2019 18:59:11 +0000 (14:59 -0400)]
ecore/timer: correctly handle recursive deletion of legacy timers

if a legacy timer callback returns false, the timer is deleted. in the
case where the legacy timer is deleted inside the callback while the same
timer is ticking recursively, however, the deletion must be deferred until
the outer-most frame of the timer's callstack has returned from the callback
in order to avoid improperly handling the timer

@fix

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10545

4 years agoevas/font: simplify handling strings when constructing font names
Mike Blumenkrantz [Thu, 17 Oct 2019 14:28:55 +0000 (10:28 -0400)]
evas/font: simplify handling strings when constructing font names

no need to strcpy here when we can just pass the length to stringshare directly

CID 1382854

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10441

4 years agoefl_ui: mark Efl.Ui.Focus.Autoscroll_Mode beta
Mike Blumenkrantz [Mon, 28 Oct 2019 12:36:07 +0000 (08:36 -0400)]
efl_ui: mark Efl.Ui.Focus.Autoscroll_Mode beta

this cannot currently be used for anything and was not explicitly stabilized

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10540

4 years agoefl_ui: remove Efl.Ui.Slider_Indicator_Visible_Mode
Mike Blumenkrantz [Mon, 28 Oct 2019 12:35:25 +0000 (08:35 -0400)]
efl_ui: remove Efl.Ui.Slider_Indicator_Visible_Mode

this cannot be used for anything and serves no purpose

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10539

4 years agoeina: introduce a simple malloc near implementation for Chained Mempool.
Cedric BAIL [Fri, 25 Oct 2019 16:11:06 +0000 (09:11 -0700)]
eina: introduce a simple malloc near implementation for Chained Mempool.

This actually seems to significantly and reliably improve speed result with
expedite in a higher way than I expected (~10%) which is neat, I guess.

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D10537

4 years agoeina: make use of the new near allocation policy for Eina_List.
Cedric BAIL [Fri, 25 Oct 2019 15:00:23 +0000 (08:00 -0700)]
eina: make use of the new near allocation policy for Eina_List.

The idea is to improve memory locality and hopefully get better cache hit
in general.

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D10536

4 years agoeina: introduce an API for requesting memory near already allocated memory from an...
Cedric BAIL [Fri, 25 Oct 2019 15:01:35 +0000 (08:01 -0700)]
eina: introduce an API for requesting memory near already allocated memory from an Eina_Mempool.

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D10535

4 years agoeina: change default policy of Chained Mempool to recycle memory right away.
Cedric BAIL [Fri, 25 Oct 2019 14:58:31 +0000 (07:58 -0700)]
eina: change default policy of Chained Mempool to recycle memory right away.

Until now, Chained Mempool would first empty its pool and the started to
recycle memory. Now it does always try to recycle first. This should limit
memory fragmentation to some extend.

Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Differential Revision: https://phab.enlightenment.org/D10534

4 years agoelm/widget: error on null params for tree_unfocusable functions
Mike Blumenkrantz [Wed, 30 Oct 2019 13:29:56 +0000 (09:29 -0400)]
elm/widget: error on null params for tree_unfocusable functions

Summary: these should error so the user can detect that they screwed up

Reviewers: devilhorns

Reviewed By: devilhorns

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10563

4 years agoedje_calc: Remove unused flag.
Woochanlee [Wed, 30 Oct 2019 12:10:09 +0000 (08:10 -0400)]
edje_calc: Remove unused flag.

Summary: This has marked remove me.

Reviewers: raster, Hermet, zmike, devilhorns

Reviewed By: devilhorns

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10561

4 years agocsharp: FunctionWrapper doc fixes
Lauro Moura [Wed, 30 Oct 2019 11:39:11 +0000 (12:39 +0100)]
csharp: FunctionWrapper doc fixes

Reviewers: felipealmeida, brunobelo, segfaultxavi, woohyun

Reviewed By: segfaultxavi

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10430

4 years agoeina: add comparison macros for Eina_Size2D and Eina_Position2D
Mike Blumenkrantz [Tue, 29 Oct 2019 18:52:58 +0000 (14:52 -0400)]
eina: add comparison macros for Eina_Size2D and Eina_Position2D

Summary:
I'm tired of typing all this out. it's exhausting.

also add a couple usages internally to verify that this works as expected

@feature

Reviewers: cedric, bu5hm4n, devilhorns

Reviewed By: devilhorns

Subscribers: devilhorns, bu5hm4n, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10557

4 years agointerfaces: replace doubles with Efl.Gfx.Align where appropriate
Mike Blumenkrantz [Tue, 29 Oct 2019 17:42:07 +0000 (18:42 +0100)]
interfaces: replace doubles with Efl.Gfx.Align where appropriate

Summary:
this makes the types more explicit
Depends on D10554

Reviewers: segfaultxavi

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10555

4 years agoefl/gfx: add Efl.Gfx.Align type
Mike Blumenkrantz [Tue, 29 Oct 2019 17:42:04 +0000 (18:42 +0100)]
efl/gfx: add Efl.Gfx.Align type

Summary:
this can be used to more explicitly specify that a double is intended to
be a value of 0.0 to 1.0 for the purpose of aligning objects. it also avoids
the need to copy and paste the same docs around everywhere

Reviewers: segfaultxavi

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10554

4 years agomeson: add embryo as a dep of edje
Marcel Hollerbach [Tue, 29 Oct 2019 16:52:07 +0000 (17:52 +0100)]
meson: add embryo as a dep of edje

if you want to build a edje theme without building all of efl, you can
do that now. Before embryo was missing and no scripts could be compiled.
Now this is possible and works.

4 years agomeson: redo evas building
Marcel Hollerbach [Tue, 29 Oct 2019 16:54:55 +0000 (12:54 -0400)]
meson: redo evas building

Summary:
before recent times we had to support static and shared building based
on the options of the user, which forced us to complicate our build with
the evas_goal hack. the evas_goal hack more or less was the idea of
"faking" the evas build in the evas directory, finish all the .eo
generation there, then build the modules and make all the static files
ready. Then build everything in evas_goal.

Now, that we just build everything the same always, we can simply build
it in the evas way (removing the evas_goal hack FINALLY), as the same modules
are build statically and shared.
This also gives us the possibility to build the shared image loaders
*again* the the modules directory, which unbreaks peoples build scripts
who packaged loader files seperatly.

Reviewers: zmike, raster, cedric, stefan_schmidt

Reviewed By: zmike

Subscribers: #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10548

4 years agoci: remove existing base_dir ccache config on osx build before adding new one
Mike Blumenkrantz [Tue, 29 Oct 2019 16:20:33 +0000 (12:20 -0400)]
ci: remove existing base_dir ccache config on osx build before adding new one

Summary: ensure that this file does not balloon to infinite size

Reviewers: bu5hm4n

Reviewed By: bu5hm4n

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10538

4 years agomono: encapsulate internal workaround members
Lauro Moura [Tue, 29 Oct 2019 14:57:55 +0000 (11:57 -0300)]
mono: encapsulate internal workaround members

Summary: Depends on D10345

Test Plan: meson setup -Dbindings=mono,cxx -Dmono-beta=true

Reviewers: segfaultxavi, Jaehyun_Cho, YOhoho

Reviewed By: YOhoho

Subscribers: cedric, #reviewers, woohyun, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10346

4 years agomono: encapsulate internal iwrapper members
Lauro Moura [Tue, 29 Oct 2019 14:56:38 +0000 (11:56 -0300)]
mono: encapsulate internal iwrapper members

Summary: Depends on D10339

Test Plan: meson setup -Dbindings=mono,cxx -Dmono-beta=true

Reviewers: segfaultxavi, Jaehyun_Cho, YOhoho

Reviewed By: YOhoho

Subscribers: cedric, #reviewers, #committers, woohyun

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10345

4 years agomono: encapsulate internal delegates
Lauro Moura [Tue, 29 Oct 2019 14:56:04 +0000 (11:56 -0300)]
mono: encapsulate internal delegates

Summary: Depends on D10337

Test Plan: meson setup -Dbindings=mono,cxx -Dmono-beta=true

Reviewers: segfaultxavi, Jaehyun_Cho, YOhoho, brunobelo

Reviewed By: YOhoho

Subscribers: cedric, #reviewers, #committers, woohyun

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D10339

4 years agoedje: selectively inhibit content,changed events when changing swallowed parts
Mike Blumenkrantz [Thu, 10 Oct 2019 17:49:45 +0000 (13:49 -0400)]
edje: selectively inhibit content,changed events when changing swallowed parts

ref T8321

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D10508

4 years agotests/genlist: add behavior testing for nearly all genlist events
Mike Blumenkrantz [Fri, 25 Oct 2019 14:06:56 +0000 (10:06 -0400)]
tests/genlist: add behavior testing for nearly all genlist events

this simulates a number of user interactions to ensure that various
functionalities of genlist actually work, including but not limited to:
* selection
* highlighting
* item focus
* scrolling
* filtering
* dragging
* item reordering

ref T8433

Differential Revision: https://phab.enlightenment.org/D10528

4 years agotests/loop_timer: add port of ecore_timer behavior tests for loop_timer
Mike Blumenkrantz [Fri, 25 Oct 2019 17:56:04 +0000 (13:56 -0400)]
tests/loop_timer: add port of ecore_timer behavior tests for loop_timer

ensure that the behavior here is also working exactly as expected

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10531

4 years agotests/ecore: make timer behavior test even more strict
Mike Blumenkrantz [Fri, 25 Oct 2019 16:50:48 +0000 (12:50 -0400)]
tests/ecore: make timer behavior test even more strict

we need to also verify that timers will process out of order solely based
on their timestamps and ignoring whether they are "recently-added"

additionally verify the behavior of timer interval changing and re-instantiating

ref T8434

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10530

4 years agoelm/genlist: fix item focus unregister on item move
Mike Blumenkrantz [Mon, 28 Oct 2019 19:45:56 +0000 (15:45 -0400)]
elm/genlist: fix item focus unregister on item move

if the block is realized, its items have been registered into the focus
manager and must be unregistered to avoid double-registering

@fix

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D10544

4 years agoelm/genlist: hide cached item contents and mark content unfocusable during calc
Mike Blumenkrantz [Mon, 28 Oct 2019 19:48:08 +0000 (15:48 -0400)]
elm/genlist: hide cached item contents and mark content unfocusable during calc

cached item contents should already be hidden by the edje clipper, so this simply
changes their visible state to break them out of the focus calcs

contents must also be explicitly marked as unfocusable during calc-only realize
operations in order to avoid triggering a full focus recalc which will error due
to missing focus adapter in the item block

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D10543

4 years agoelm/genlist: set pan need recalc during item move on item position change
Mike Blumenkrantz [Fri, 25 Oct 2019 19:05:29 +0000 (15:05 -0400)]
elm/genlist: set pan need recalc during item move on item position change

item move operations require pan recalc in order to process the item block
positioning updates

@fix

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D10533

4 years agoelm/genlist: fix "drag" smart callback
Mike Blumenkrantz [Thu, 24 Oct 2019 15:36:30 +0000 (11:36 -0400)]
elm/genlist: fix "drag" smart callback

this is only a smart callback and not an eo callback

@fix

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10527

4 years agoelm/genlist: trigger filter,done callback immediately if no filter queue exists
Mike Blumenkrantz [Wed, 23 Oct 2019 14:38:53 +0000 (10:38 -0400)]
elm/genlist: trigger filter,done callback immediately if no filter queue exists

the documentation says this should trigger when filtering is done, and if
no filtering is pending then it is done

@fix

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10526

4 years agoelm/genlist: slightly simplify use of _item_filtered_get()
Mike Blumenkrantz [Wed, 23 Oct 2019 14:36:01 +0000 (10:36 -0400)]
elm/genlist: slightly simplify use of _item_filtered_get()

passing the smart data here (which we always have) makes the function
a little clearer to read. similarly, we can check whether the filter_data
pointer is set inside the function to avoid having to check it everywhere else

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10525

4 years agoelm/genlist: defer recalc when applying a name filter
Mike Blumenkrantz [Wed, 23 Oct 2019 14:27:04 +0000 (10:27 -0400)]
elm/genlist: defer recalc when applying a name filter

this may be called successively during the same mainloop iteration,
so it's important to defer this as much as possible

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10524

4 years agoelm/genlist: remove calc jobs
Mike Blumenkrantz [Mon, 21 Oct 2019 20:05:33 +0000 (16:05 -0400)]
elm/genlist: remove calc jobs

now that we have a sane and consistent sizing calc mechanism, we no longer
need to be randomly creating jobs to do additional side calcs in addition
to our other multiple bespoke calculation methods

instead, we can now call the calc function directly during the group calc
to perform all the calcs at once and avoid the overhead of constantly triggering
calc jobs

a possible future improvement here could be to remove the _calc_job() call in
the internal pan object's group_calculate, as this is likely a duplicated op
but it will require additional testing to verify

unit test performance (e.g., tree) increases roughly 50-80% after this patch

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10523

4 years agoelm/genlist: don't process entire item queue on each item add
Mike Blumenkrantz [Mon, 21 Oct 2019 20:03:56 +0000 (16:03 -0400)]
elm/genlist: don't process entire item queue on each item add

this is a colossal waste of time. it ends up realizing every single item
immediately even if it won't be visible, which defeats the purpose of all
the batching and viewport calculations that genlist explicitly does to
avoid exactly this type of behavior

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10522

4 years agoelm/genlist: freeze/thaw canvas when processing item queue
Mike Blumenkrantz [Mon, 21 Oct 2019 20:01:34 +0000 (16:01 -0400)]
elm/genlist: freeze/thaw canvas when processing item queue

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10521

4 years agoelm/genlist: only add calc job on theme apply if already finalized
Mike Blumenkrantz [Mon, 21 Oct 2019 19:53:35 +0000 (15:53 -0400)]
elm/genlist: only add calc job on theme apply if already finalized

not a required operation otherwise

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10519

4 years agoelm/genlist: remove misleading "changed" signal
Mike Blumenkrantz [Mon, 21 Oct 2019 18:02:34 +0000 (14:02 -0400)]
elm/genlist: remove misleading "changed" signal

this was never implemented or documented anywhere and serves only to be
misleading for anyone reading the code

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10518

4 years agoelm/genlist: fix item highlight/unhighlight eventing on clicks
Mike Blumenkrantz [Mon, 21 Oct 2019 15:02:18 +0000 (11:02 -0400)]
elm/genlist: fix item highlight/unhighlight eventing on clicks

* highlight should only be triggered on mouse down, as that is the point of
  interaction with an item which indicates it is "in use"; a mouse-up event
  cannot occur on an item which has not previously received a mouse-down event,
  so toggling this on mouse-up will be wrong/duplicated 100% of the time
* unhighlight should only be triggered during mouse-up events if the list is
  in no-select mode, as it will otherwise be implicitly during selection if
  necessary

this should ensure that these events are emitted exactly one time and correctly
for each click event

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10517

4 years agotests/ecore: add more grueling timer throughput test
Mike Blumenkrantz [Thu, 24 Oct 2019 18:26:24 +0000 (14:26 -0400)]
tests/ecore: add more grueling timer throughput test

this verifies that:
* newly-created timers are not triggered in the next loop iteration
* newly-created timers can be triggered the second loop after created
* multiple timers can be triggered in a single loop iteration
* timers are effectively added to the pending timer list

ref T8434

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D10516