platform/upstream/efl.git
5 years agocanvas vg: fix wrong key pass for caching surface. 43/202643/2
Hermet Park [Tue, 2 Apr 2019 06:17:24 +0000 (15:17 +0900)]
canvas vg: fix wrong key pass for caching surface.

There is a wrong case that fails caching ector surfaces.
Those vector objects manually constructing shapes,
passed wrong pointer as the cacahing key instead of root node pointer.

This fixes it.

Change-Id: I5f547e896d1d380b2b7fbad00587609f54ce6ec2

5 years agospec: fix a dependencies for devel package properly 22/200922/2
Wonki Kim [Wed, 6 Mar 2019 04:45:38 +0000 (13:45 +0900)]
spec: fix a dependencies for devel package properly

package descriptions for edje-devel, elementary-devel was not described properly
they need libraries itself. so that this patch fix it.

Change-Id: I8daa2490b91546ca1083a2adfa69e0e7a02a55e6
Signed-off-by: Wonki Kim <wonki_.kim@samsung.com>
5 years agospec: exclude unused binaries. 89/202589/2
Wonki Kim [Mon, 1 Apr 2019 07:56:05 +0000 (16:56 +0900)]
spec: exclude unused binaries.

this pathc excludes some binaries from the packiging list

Change-Id: Ic967ad3846ce60d0c53b3c85296b0261dbe71be3
Signed-off-by: Wonki Kim <wonki_.kim@samsung.com>
5 years agomeson: fix a meson option to install eo files 55/202655/2 submit/tizen/20190402.080200
Wonki Kim [Tue, 2 Apr 2019 07:50:21 +0000 (16:50 +0900)]
meson: fix a meson option to install eo files

as before, eo files are installed as default,
however, after some commit, it doesn't for now.

this patch fixes a meson option to install those files.

Change-Id: I1060ba5476533f479a18b86df020a7e051792cdb

5 years agomigratino: remove beta tags 50/202650/2
JunsuChoi [Tue, 2 Apr 2019 07:39:56 +0000 (16:39 +0900)]
migratino: remove beta tags

@tizen_fix

Change-Id: If2b152273cc9d3fa31a8eac59810e4db7f7cdad5

5 years agoelm_widget_item_eo.legacy: change @ingroup to Elm_General 44/202644/1
Jaehyun Cho [Tue, 2 Apr 2019 06:33:04 +0000 (15:33 +0900)]
elm_widget_item_eo.legacy: change @ingroup to Elm_General

The @ingroup of elm_object_item APIs is Elm_General in elm_object_item.h.

Since legacy prefix of Elm_Widget_Item is elm_object_item,
the prefix of Elm_Widget_Item APIs is elm_object_item.

Consequently, to display the legacy Elm_Widget_Item APIs with
elm_object_item APIs, the @ingroup is changed from Elm_Object_Item_Group
to Elm_General.

Change-Id: Ic023a57f15ee3968fb0a5cfa09d9ab33c4adf420

5 years agoecore-drm2: Add missing @ingroup for some doxy submit/tizen/20190402.051338
Christopher Michael [Mon, 1 Apr 2019 14:31:08 +0000 (10:31 -0400)]
ecore-drm2: Add missing @ingroup for some doxy

Small patch to add missing @ingroup for doxygen comments, and correct
one that was in the wrong group

@fix

5 years agoefl_ui_selection_manager: Don't leak malloc'd data
Christopher Michael [Mon, 1 Apr 2019 14:02:57 +0000 (10:02 -0400)]
efl_ui_selection_manager: Don't leak malloc'd data

Summary:
Coverity reports that we potentially leak char *s here. If we do not
have 'data_ret', then the malloc'd 's' sould be freed as we are not
going to use it.

Fixes Coverity CID1396949

@fix

Reviewers: raster, cedric, bu5hm4n, zmike

Reviewed By: bu5hm4n

Subscribers: #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8523

5 years agoevas-font-dir: Minor formatting fixes
Christopher Michael [Mon, 1 Apr 2019 13:05:57 +0000 (09:05 -0400)]
evas-font-dir: Minor formatting fixes

NB: No functional changes

5 years agoelm - fix harmless warning for clean build
Carsten Haitzler (Rasterman) [Sat, 30 Mar 2019 16:54:19 +0000 (16:54 +0000)]
elm - fix harmless warning for clean build

5 years agoelm - fix harmless warning for clean build
Carsten Haitzler (Rasterman) [Sat, 30 Mar 2019 16:49:48 +0000 (16:49 +0000)]
elm - fix harmless warning for clean build

5 years agoelm - fix harmless warning for clean build
Carsten Haitzler (Rasterman) [Sat, 30 Mar 2019 16:49:21 +0000 (16:49 +0000)]
elm - fix harmless warning for clean build

5 years agoefl_ui_table_layout: calculate cell size with colspan, rowspan property
Yeongjong Lee [Sat, 30 Mar 2019 08:49:57 +0000 (08:49 +0000)]
efl_ui_table_layout: calculate cell size with colspan, rowspan property

We should consider occupied cells by colspan, rowspan property.

ref T7753

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8484

5 years agobuild: improve autotools generation of elm config
Mike Blumenkrantz [Thu, 28 Mar 2019 15:25:21 +0000 (11:25 -0400)]
build: improve autotools generation of elm config

 - don't generate and re-link on every make rule
 - fix distcheck

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Reviewed-by: YeongJong Lee <yj34.lee@samsung.com>
Differential Revision: https://phab.enlightenment.org/D8506

5 years agoREADME.meson: fix typo + rewrap
Boris Faure [Fri, 29 Mar 2019 19:21:55 +0000 (20:21 +0100)]
README.meson: fix typo + rewrap

5 years agotests: abort on errors during genlist expand test, not warnings
Mike Blumenkrantz [Fri, 29 Mar 2019 17:58:14 +0000 (18:58 +0100)]
tests: abort on errors during genlist expand test, not warnings

Summary:
log level=2 is the warning level, which is not super useful since
there's currently billions of eo warnings occuring in every function
call

Reviewers: cedric, segfaultxavi

Reviewed By: cedric, segfaultxavi

Subscribers: segfaultxavi, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8515

5 years agoeina_log: reset logging callback to default when null is set as the callback
Mike Blumenkrantz [Fri, 29 Mar 2019 17:50:30 +0000 (18:50 +0100)]
eina_log: reset logging callback to default when null is set as the callback

Summary:
passing null here causes any log message to crash the app and is probably not
the intended result

Reviewers: cedric, segfaultxavi

Reviewed By: cedric, segfaultxavi

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8513

5 years agoelm_entry: Speedup finding new line, prevent readind invalid memory
Ali Alzyod [Fri, 29 Mar 2019 13:52:51 +0000 (09:52 -0400)]
elm_entry: Speedup finding new line, prevent readind invalid memory

Summary:
1- Speed up detecting new lines.
```
if (!strncmp(text, "<br", 3) || !strncmp(text, "<ps", 3))
```
This will cause 6 comparisons (if one of conditions did not meet), or at least 3 comparisons.

this is changed to
```
if (!strncmp(text, "<", 1))
```

2- Speedup detecting lines

If this condition is true, we should increment the string for next iteration 3 times, not just one
```
if (!strncmp(text, "<br", 3) || !strncmp(text, "<ps", 3))
```

if '<' founded then 'pr' or 'br', we will skip 3 characters for next iteration.

```
if (!strncmp(text, "<", 1))
      {
         text++;
         len--;
         if (!strncmp(text, "br", 2) || !strncmp(text, "ps", 2))
         {
            text += 2;
            len -= 2;
```

3- Prevent reading invalid memory out of the string

```
if (text[3] == '>' || ((text[3] == '/') && (text[4] == '>')))
```
string could reach last char in string (original string ends with "<br")

but now we will check if remaining string length allow comparison :

```
if (text[0] == '>' || (len > 1 && ((text[0] == '/') && (text[1] == '>'))))
```

Test Plan:
```

static int

oldFunc(const char *text)
{
    if (!text)
        return 0;

    while (*text)
    {
        if (!strncmp(text, "<br", 3) || !strncmp(text, "<ps", 3))
        {
            if (text[3] == '>' || ((text[3] == '/') && (text[4] == '>')))
            {
                return 1;
            }
        }
        text++;
    }
    return 0;
}

static int
newFunc(const char *text)
{
    if (!text)
        return 0;
    char *pTemp = (char *)text;

    while (pTemp = strchr(pTemp, '<'))
    {
        pTemp++;
        if (!strncmp(pTemp, "br", 2) || !strncmp(pTemp, "ps", 2))
        {
            pTemp += 2;
            if (pTemp[0] != '\0' && (pTemp[0] == '>' || (pTemp[0] == '/' && pTemp[1] == '>')))
            {
                return 1;
            }
        }
    }

    return 0;
}

int main()
{

    int counter = 1000;
    srand(time(NULL));
    char pStr[50001] = {0};
    char AllChars[] = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789<>";

    int AllCharsLen = strlen(AllChars);
    for (int i = 0; i < 50000; i++)
        pStr[i] = AllChars[rand() % AllCharsLen];

    clock_t start, end;
    double total_Time1 = 0;
    int i;

    for (int j = 0; j < 3; j++)
    {
        if (j == 0)
        {
            printf("random String\n");
        }
        else if (j == 1)
        {
            printf("With Random <br/>\n");
            int location = rand()%(5000 - 5);
            pStr[location++] = '<';
            pStr[location++] = 'b';
            pStr[location++] = 'r';
            pStr[location++] = '/';
            pStr[location++] = '>';
        }
        else if (j == 2)
        {
            printf("With Random <ps>\n");
            int location = rand()%(5000 - 4);
            pStr[location++] = '<';
            pStr[location++] = 'p';
            pStr[location++] = 's';
            pStr[location++] = '>';
        }

        start = clock();
        for (i = 0; i < counter; i++)
            oldFunc(pStr);
        end = clock();
        total_Time1 = ((double)(end - start)) / CLOCKS_PER_SEC;
        printf("original = %f has new Line = %i\n", total_Time1, oldFunc(pStr));

        start = clock();
        for (i = 0; i < counter; i++)
            newFunc(pStr);
        end = clock();
        total_Time1 = ((double)(end - start)) / CLOCKS_PER_SEC;
        printf("modified = %f has new line = %i\n\n", total_Time1, newFunc(pStr));
    }
}
```

output:

random String
original = 2.523000 has new Line = 0
modified = 0.090000 has new line = 0

With Random <br/>
original = 0.081000 has new Line = 1
modified = 0.003000 has new line = 1

With Random <ps>
original = 0.016000 has new Line = 1
modified = 0.001000 has new line = 1

Reviewers: zmike, woohyun, bowonryu

Reviewed By: zmike

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

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8497

5 years agodocs: Fix build break because of missed EO import
Xavi Artigas [Fri, 29 Mar 2019 13:12:37 +0000 (14:12 +0100)]
docs: Fix build break because of missed EO import

5 years agodocs: Use Eina.Value references en efl_ui_format docs
Xavi Artigas [Fri, 29 Mar 2019 13:08:22 +0000 (14:08 +0100)]
docs: Use Eina.Value references en efl_ui_format docs

5 years agoefl_ui_layout: ensure that resize_obj is present before emitting signals
Marcel Hollerbach [Fri, 29 Mar 2019 12:25:54 +0000 (13:25 +0100)]
efl_ui_layout: ensure that resize_obj is present before emitting signals

otherwise there will be warnings because of calling api on NULL objects.

Differential Revision: https://phab.enlightenment.org/D8512

5 years agoefl_ui_widget: move from elm_widget_top_get to provider_find
Marcel Hollerbach [Tue, 26 Mar 2019 09:15:15 +0000 (10:15 +0100)]
efl_ui_widget: move from elm_widget_top_get to provider_find

this resolves a lot of cases where focus_highlight API was called on a
object, which is not a efl_ui_win object. With this patch we ensure that
the object is always a window.

Reviewed-by: Xavi Artigas <xavierartigas@yahoo.es>
Differential Revision: https://phab.enlightenment.org/D8476

5 years agoefl_ui_widget: add implementation for finding the window
Marcel Hollerbach [Tue, 26 Mar 2019 08:59:16 +0000 (09:59 +0100)]
efl_ui_widget: add implementation for finding the window

the problem with the previous implementation (just redirect the calls to
the widget_parent then to the efl_parent is that after invalidate its
impossible to find the window where the widget is in. However, there are
cases where we want to have access to the window of the widget, for
example, to invalidate focus highlight etc..
The window of a widget is always constant, and cannot be changed (as the
evas object cannot hop accross different evas)

Reviewed-by: Jaehyun Cho <jae_hyun.cho@samsung.com>
Differential Revision: https://phab.enlightenment.org/D8475

5 years agodocs: Efl.Ui.Layout_Base update theme docs
Xavi Artigas [Thu, 28 Mar 2019 16:12:27 +0000 (16:12 +0000)]
docs: Efl.Ui.Layout_Base update theme docs

If docs are present at property and set/get levels only one is used.

Ref T7717

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8494

5 years agodocs: Clarify Efl.Ui.Win exit_on_close methods
Xavi Artigas [Fri, 29 Mar 2019 09:30:44 +0000 (09:30 +0000)]
docs: Clarify Efl.Ui.Win exit_on_close methods

exit_on_close and exit_on_all_windows_closed deserve a bit of clarification
since they have very similar meanings.
Also, add proper Eina.Value doc references.
Ref T7717

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8511

5 years agodocs: Add Eina.Value extern to eina_types.eot
Xavi Artigas [Fri, 29 Mar 2019 09:25:02 +0000 (09:25 +0000)]
docs: Add Eina.Value extern to eina_types.eot

Eina.Values are built-in eolian types, accessed through any_value and
any_value_ptr. However, these types cannot be used in doc references.
Adding a placeholder extern struct Eina.Value causes no harm, and will allow
referencing the type from EO docs later on.
In C#, Eina.Value is defined in the manual binding code so the doc reference
resolves to a valid type.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8510

5 years agoci: check correct test log for meson build
Mike Blumenkrantz [Thu, 28 Mar 2019 22:04:55 +0000 (18:04 -0400)]
ci: check correct test log for meson build

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8504

5 years agoci: use meson test runner with ninja build and use dbus in tests
Mike Blumenkrantz [Thu, 28 Mar 2019 15:26:55 +0000 (11:26 -0400)]
ci: use meson test runner with ninja build and use dbus in tests

this fixes dbus usage in tests

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8503

5 years agotest/efl_ui_pager: fix demo
Jaeun Choi [Fri, 29 Mar 2019 11:03:16 +0000 (20:03 +0900)]
test/efl_ui_pager: fix demo

- use unpack_at function than unpack function
- add CLEAR option

5 years agotest/efl_ui_pager_scroll: fix demo
Jaeun Choi [Tue, 19 Feb 2019 09:42:31 +0000 (18:42 +0900)]
test/efl_ui_pager_scroll: fix demo

- use radio than check for loop mode
- use unpack_at function than unpack function
- add CLEAR option

5 years agoefl_ui_pager: implement unpack_all/clear function
Jaeun Choi [Fri, 29 Mar 2019 10:48:58 +0000 (19:48 +0900)]
efl_ui_pager: implement unpack_all/clear function

5 years agoefl_ui_pager: return mouse_down function if cnt equals zero
Jaeun Choi [Fri, 29 Mar 2019 10:23:20 +0000 (19:23 +0900)]
efl_ui_pager: return mouse_down function if cnt equals zero

5 years agoefl_ui_pager: implement unpack_at function
Jaeun Choi [Thu, 28 Mar 2019 10:26:01 +0000 (19:26 +0900)]
efl_ui_pager: implement unpack_at function

5 years agoefl_ui_pager: refactor unpack function
Jaeun Choi [Thu, 28 Mar 2019 10:21:21 +0000 (19:21 +0900)]
efl_ui_pager: refactor unpack function

5 years agoefl_ui_pager: add a missing condition
Jaeun Choi [Fri, 22 Mar 2019 08:51:17 +0000 (17:51 +0900)]
efl_ui_pager: add a missing condition

5 years agoefl_ui_pager: fix pack function
Jaeun Choi [Fri, 22 Mar 2019 02:30:34 +0000 (11:30 +0900)]
efl_ui_pager: fix pack function

5 years agoelm_priv.h: remove unnecesary line
Jaeun Choi [Tue, 19 Mar 2019 05:32:18 +0000 (14:32 +0900)]
elm_priv.h: remove unnecesary line

5 years agoefl_ui_pager: disable loop if items are not enough after unpacking
Jaeun Choi [Tue, 19 Feb 2019 09:44:26 +0000 (18:44 +0900)]
efl_ui_pager: disable loop if items are not enough after unpacking

5 years agomono-docs: Minor ammendments to Eina.Value
Xavi Artigas [Fri, 29 Mar 2019 08:51:56 +0000 (09:51 +0100)]
mono-docs: Minor ammendments to Eina.Value

5 years agoefl_ui_widget: reintroduce legacy behaviour
Marcel Hollerbach [Sun, 24 Mar 2019 17:58:50 +0000 (18:58 +0100)]
efl_ui_widget: reintroduce legacy behaviour

before the refactoring of the disabled property, there was no way to
enable a widget which has a disabled tree. This here however enables
this to work again like this. The user will be told with an error
message. The integraty of the property is maintained accross reparents.

Reviewed-by: Jaehyun Cho <jae_hyun.cho@samsung.com>
Differential Revision: https://phab.enlightenment.org/D8459

5 years agoefl_ui_widget: add tests for parent and disalbed property
Marcel Hollerbach [Sun, 24 Mar 2019 17:22:15 +0000 (18:22 +0100)]
efl_ui_widget: add tests for parent and disalbed property

this just adds more coverage over the behaviour of efl_ui_widget
properties.

Reviewed-by: Jaehyun Cho <jae_hyun.cho@samsung.com>
Differential Revision: https://phab.enlightenment.org/D8458

5 years agoefl_ui_test_widget: ensures tests do not error
Marcel Hollerbach [Sun, 24 Mar 2019 17:22:03 +0000 (18:22 +0100)]
efl_ui_test_widget: ensures tests do not error

Reviewed-by: Jaehyun Cho <jae_hyun.cho@samsung.com>
Differential Revision: https://phab.enlightenment.org/D8457

5 years agoRevert "evas gl engines - avoid getting context if possible"
Carsten Haitzler (Rasterman) [Fri, 29 Mar 2019 00:16:22 +0000 (00:16 +0000)]
Revert "evas gl engines - avoid getting context if possible"

This reverts commit e7771438a25382a2e4fabe33753456dcde8cab40.

This should fix T7764

zmike - you're right. thanks for narrowing down the commit... revert
time.

5 years agoremove the definition of HAVE_WASAPI as it is never used
Vincent Torri [Thu, 28 Mar 2019 13:23:20 +0000 (09:23 -0400)]
remove the definition of HAVE_WASAPI as it is never used

Summary: HAVE_WASAPI is never used

Test Plan: compilation

Reviewers: zmike, cedric, raster

Reviewed By: zmike

Subscribers: #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8493

5 years agoefl_ui_alert_popup: add new feature for applying side button style of alert popup
Taehyub Kim [Thu, 28 Mar 2019 09:48:03 +0000 (18:48 +0900)]
efl_ui_alert_popup: add new feature for applying side button style of alert popup

Summary: This feature will apply side button style for each left and right button of alert popup

Reviewers: Jaehyun_Cho, cedric

Reviewed By: Jaehyun_Cho

Subscribers: #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8492

5 years agoelm_entry: CRI if efl_file methods are called directly on this object
Mike Blumenkrantz [Wed, 27 Mar 2019 15:25:28 +0000 (11:25 -0400)]
elm_entry: CRI if efl_file methods are called directly on this object

eo methods should not be called on legacy objects

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8487

5 years agodocs: Document event info calling convention
Xavi Artigas [Wed, 27 Mar 2019 15:10:57 +0000 (15:10 +0000)]
docs: Document event info calling convention

Both at the emitter (efl_event_callback_call) and the receiver
(info field in the Efl.Event structure).
The Events tutorial should repeat this.
Fixes T7760

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8486

5 years agoefl_ui_slider: block scrolling when on slider
Marcel Hollerbach [Sun, 24 Mar 2019 12:24:47 +0000 (13:24 +0100)]
efl_ui_slider: block scrolling when on slider

when a mouse cursor is over a slider, the mouse wheel should be used to
affect the state of the slider, not also the one of the slider.

ref T2529

Reviewed-by: Bowon Ryu <bowon.ryu@samsung.com>
Reviewed-by: Jaehyun Cho <jae_hyun.cho@samsung.com>
Differential Revision: https://phab.enlightenment.org/D8455

5 years agoelementary: remove some asynchronous behavior from the fileselector.
Cedric BAIL [Wed, 27 Mar 2019 22:19:32 +0000 (15:19 -0700)]
elementary: remove some asynchronous behavior from the fileselector.

Marcel notted that when using the LIST view of the fileselector on a
big directory, we end up having to wait for the entire genlist to be
populated to be able to switch to another directory. This is actually
a side effect of the populate code being triggered through an idler.
This idler was useful when the list was populated directly, but now
that we rely on Efl.Io.Model, we should be asynchronous enough that
it shouldn't be a problem to actually not be asynchronous here. By
removing the reliance on the idler, we are not queued after all the
idler and can properly short circuit all of that.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8491

5 years agoelementary: restore quick exit from wait loop in fileselector test.
Cedric BAIL [Wed, 27 Mar 2019 21:24:57 +0000 (14:24 -0700)]
elementary: restore quick exit from wait loop in fileselector test.

The test was not expecting both callback to be set when the wait loop
was started. By moving them around, it fixes the test case to only have
one relevant callback set at a time.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8490

5 years agoelc_filselector: solve the mysterium of sometimes failing tests
Marcel Hollerbach [Wed, 27 Mar 2019 20:15:53 +0000 (21:15 +0100)]
elc_filselector: solve the mysterium of sometimes failing tests

so after a phone call, two days of debugging, tears, crying etc. etc. we
finally came to a point of enlightenment! *Someone* (bu5hm4n) moved
gengrid and genlist events from eo back to smart events, so we can work
arround legacy borks and event-name collisions, at this point he did not
knew that some widgets (fileselector) already relied on those *lovely*
events. Hence this broke theoretically the testsuite, however, the
fileselector testsuite is ultimatily buggy, and the wait function does
not return false when it timeouts, (i don't know why not). So this break
was never discovered.

Additionally there is a second issue. it appears, that when we
immidiatly quit the mainloop after we have got the selected callback,
that then genlist decides to forget about the sd->selected pointer, and
NULLs that one out. Which then results in the fact that
elm_fileselector_selected_get ends up returning invalid paths.

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D8488

5 years agoelementary: make sure that the model parent being used is always the fileselector.
Cedric BAIL [Fri, 22 Mar 2019 18:20:51 +0000 (11:20 -0700)]
elementary: make sure that the model parent being used is always the fileselector.

Model provided by an item selection would have there parent being the current
model of the fileselector. Once that one is replaced by the item model, it would
automatically invalidate the model and break any further request. This lead to
a bug where you could only get into one directory before everything else
being empty.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8450

5 years agoelementary: prevent asynchronous properties change to believe target is ready when...
Cedric BAIL [Fri, 22 Mar 2019 17:31:52 +0000 (10:31 -0700)]
elementary: prevent asynchronous properties change to believe target is ready when it is not in fileselector.

In some case, the properties changed event would be triggered first on the object model
instead of the target model. This now enforce that the target will be the first model
to handle and react on the information.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8449

5 years agoelementary: fix fileselector entry support to define path manually.
Cedric BAIL [Thu, 21 Mar 2019 22:04:46 +0000 (15:04 -0700)]
elementary: fix fileselector entry support to define path manually.

There was no need in the first place to do all this asynchronous work here.
It is more robust to do it in sync.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8448

5 years agoeio: work around the lack of integration between Ecore_Thread and Eina_Future.
Cedric BAIL [Fri, 15 Mar 2019 21:38:24 +0000 (14:38 -0700)]
eio: work around the lack of integration between Ecore_Thread and Eina_Future.

Ecore_Thread excpect resolution of the error and done case to be instantaneous,
while Eina_Future default scheduler linked with Ecore main loop is build around
asynchronous answer. This create a lot of potential. A better patch would be
to provide an Ecore_Thread helper that does the integration properly. Sadly
we are in release now, so this is basically what an helper would do, but
contained inside Efl_Io_Manager.

This also solve the same problem as D7970 and D8053, but it should avoid its
side effect.

Reviewed-by: YeongJong Lee <yj34.lee@samsung.com>
Differential Revision: https://phab.enlightenment.org/D8371

5 years agoeio: add test to ensure proper lifecycle of Efl_Io object and futures.
Yeongjong Lee [Fri, 15 Mar 2019 21:51:54 +0000 (14:51 -0700)]
eio: add test to ensure proper lifecycle of Efl_Io object and futures.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8372

5 years agoeio: Efl.Io.Model should not make request when the object is invalidating itself.
Cedric BAIL [Fri, 15 Mar 2019 22:54:26 +0000 (15:54 -0700)]
eio: Efl.Io.Model should not make request when the object is invalidating itself.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8373

5 years agoeio: remove unecessary use of weak reference.
Cedric BAIL [Fri, 15 Mar 2019 23:55:07 +0000 (16:55 -0700)]
eio: remove unecessary use of weak reference.

There is no point in keeping a pointer to the main loop now that we
are using efl_future_then. This resolve potential bug with leftover
dangling weak reference as efl_future_then do require a free case
otherwise.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8374

5 years agoelementary: destroy fileselector children when they are not itemized yet.
Cedric BAIL [Fri, 15 Mar 2019 23:56:33 +0000 (16:56 -0700)]
elementary: destroy fileselector children when they are not itemized yet.

As we now do everything asynchronously, we do have model representing child
of the main model that don't provide enough information to be displayed yet.
This are not tracked by a genlist item, nor are they a child of the
fileselector. To properly handle their lifecycle, it is necessary to unref
them manually explicitely.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8375

5 years agoeio: guarantee that we will at least process one request per loop iteration for very...
Cedric BAIL [Thu, 21 Mar 2019 20:33:30 +0000 (13:33 -0700)]
eio: guarantee that we will at least process one request per loop iteration for very slow system.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8447

5 years agoelementary: improve lifecycle of model object in the fileselector widget.
Cedric BAIL [Wed, 13 Mar 2019 21:54:39 +0000 (14:54 -0700)]
elementary: improve lifecycle of model object in the fileselector widget.

This is a minimal change and it would be best to refactor the code completely
using all the infrastructure we have now instead of the organically grown code,
but I am afraid of doing such a big change at this point of our release cycle.

Part of the improvement are use of efl_replace to make sure Eo object reference
are set to NULL once reference are dropped. Handling the case when a processed
child is actually pointing to an error. Also it is not supported by model to
get their parent stolen, so this has been fixed too. Finally setting the path
asynchronously was creating more trouble than needed, when it could be done in
a synchronous way.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8336

5 years agoeo: refactor auto_unref logic used by efl_part.
Cedric BAIL [Thu, 14 Mar 2019 21:14:53 +0000 (14:14 -0700)]
eo: refactor auto_unref logic used by efl_part.

This bring no functional change to Eo and efl_part.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8357

5 years agoeo: enforce auto_unref logic at the end of efl_unref execution.
Cedric BAIL [Thu, 14 Mar 2019 21:15:28 +0000 (14:15 -0700)]
eo: enforce auto_unref logic at the end of efl_unref execution.

This allow for the safe use of efl_ref/efl_unref around an efl_part
without calling any function on that part.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8358

5 years agoRevert "efl_ui_layout: call efl_del instead of efl_unref when text part does not...
Cedric BAIL [Wed, 20 Mar 2019 17:55:29 +0000 (10:55 -0700)]
Revert "efl_ui_layout: call efl_del instead of efl_unref when text part does not exist"

This reverts commit cced5487c83c8f75e2ca40969f5749d5e6570228.

This patch was pushed just to silence warning without fixing the problem and doing
something that was incorrect (coupling an efl_del with an efl_ref).

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8429

5 years agoelementary: only apply text when the object is not invalidated and dying.
Cedric BAIL [Thu, 14 Mar 2019 21:16:50 +0000 (14:16 -0700)]
elementary: only apply text when the object is not invalidated and dying.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8359

5 years agoelementary: split the smart callback event name from the Eo name when needed.
Cedric BAIL [Wed, 13 Mar 2019 17:39:49 +0000 (10:39 -0700)]
elementary: split the smart callback event name from the Eo name when needed.

This fix the borkage of the "selected" smart event not being triggered.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8332

5 years agoeina: prevent double cancel of ongoing dispatched future.
Cedric BAIL [Wed, 13 Mar 2019 21:49:50 +0000 (14:49 -0700)]
eina: prevent double cancel of ongoing dispatched future.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8333

5 years agoeio: rely on efl_future_then to properly protect Eo object during the lifecycle of...
Cedric BAIL [Wed, 13 Mar 2019 21:50:48 +0000 (14:50 -0700)]
eio: rely on efl_future_then to properly protect Eo object during the lifecycle of the future callback.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8334

5 years agoeio: path and filename property should always remind accessible even in case of error.
Cedric BAIL [Wed, 13 Mar 2019 21:51:20 +0000 (14:51 -0700)]
eio: path and filename property should always remind accessible even in case of error.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8335

5 years agobuild: add autotools for elm config embedding
Mike Blumenkrantz [Wed, 27 Mar 2019 21:23:19 +0000 (17:23 -0400)]
build: add autotools for elm config embedding

Summary: Depends on D8481

Reviewers: bu5hm4n, devilhorns

Reviewed By: bu5hm4n

Subscribers: cedric, #reviewers, #committers

Tags: #efl_build

Differential Revision: https://phab.enlightenment.org/D8489

5 years agoelementary: embed standard config as fallback
Marcel Hollerbach [Wed, 27 Mar 2019 21:23:13 +0000 (17:23 -0400)]
elementary: embed standard config as fallback

Summary:
as you can see in T7620, there is quite some critics about how we are
handling configs in elm. This patch enables the compile-time standard
config to be used instead of some weak-wrong-usage hardcoded structure
defines somewhere in elm.

This means, that every update to the default theme will be also in the
next build embedded, without any config files installed at all, the
standard config for desktops will work, and a error will be printed, so
the user does have the possibility to interact with elm as he wishes,
while he sees this error.

fixes T7620

Reviewers: zmike, cedric, segfaultxavi, devilhorns

Reviewed By: zmike

Subscribers: devilhorns, #reviewers, #committers

Tags: #efl

Maniphest Tasks: T7620

Differential Revision: https://phab.enlightenment.org/D8481

5 years agobuild: fix meson vnc-server option text
Mike Blumenkrantz [Wed, 27 Mar 2019 18:34:48 +0000 (14:34 -0400)]
build: fix meson vnc-server option text

Summary: thx @vtorri for reporting

Reviewers: devilhorns, vtorri

Reviewed By: vtorri

Subscribers: cedric, #reviewers, vtorri, #committers

Tags: #efl_build

Differential Revision: https://phab.enlightenment.org/D8480

5 years agobuild: fix meson pixman dependency name
Mike Blumenkrantz [Wed, 27 Mar 2019 18:34:43 +0000 (14:34 -0400)]
build: fix meson pixman dependency name

Summary: thx @vtorri for reporting

Reviewers: devilhorns, vtorri

Reviewed By: vtorri

Subscribers: cedric, #reviewers, vtorri, #committers

Tags: #efl_build

Differential Revision: https://phab.enlightenment.org/D8479

5 years agoeina benchmark: fix warnings on Windows 64 bits
Vincent Torri [Wed, 27 Mar 2019 14:59:31 +0000 (10:59 -0400)]
eina benchmark: fix warnings on Windows 64 bits

Summary: long is always a 32 bits type on Windows

Test Plan: compilation

Reviewers: raster, zmike, cedric

Reviewed By: zmike

Subscribers: #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8482

5 years agotest: add a test case for elm_entry
Wonki Kim [Wed, 27 Mar 2019 14:18:41 +0000 (10:18 -0400)]
test: add a test case for elm_entry

Summary:
behaviors of elm_entry has been changed
so that this patch provides usages to keep as a test case.

Reviewers: zmike, Hermet, YOhoho

Reviewed By: zmike

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8483

5 years agoentry: move a point to do 'auto_save' to another place
Wonki Kim [Wed, 27 Mar 2019 14:18:33 +0000 (10:18 -0400)]
entry: move a point to do 'auto_save' to another place

Summary:
By reworking on efl_file, logic flow for entry has been changed.
and it causes autosave making a file that is passed to elm_entry_file_set empty.

Test Plan:
1. call elm_entry_file_set for a file.
2. check the file is not empty after calling the function.

Reviewers: zmike, bu5hm4n

Reviewed By: zmike

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8362

Change-Id: Icb8202fb9cf55401c26c628358fa39e2947b2f7b

5 years agoevas_render: Process deferred callback in the sync render case.
Woochanlee [Wed, 27 Mar 2019 04:11:15 +0000 (13:11 +0900)]
evas_render: Process deferred callback in the sync render case.

Summary:
The EVAS_CALLBACK_RENDER_POST callback has been deferred when the callback is registered during the render(inside_post_render flag on).

In the sync render case, the logic to call deferred callbacks is missing, and callbacks are not being called in certain cases.

@fix

Reviewers: ManMower, Hermet

Reviewed By: Hermet

Subscribers: zmike, cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8478

5 years agoefl.spec : add --enable-install-eo-files option
Yeongjong Lee [Wed, 27 Mar 2019 07:37:03 +0000 (16:37 +0900)]
efl.spec : add --enable-install-eo-files option

Becuase tizen use eo beta apis, it is needed to install eo files.

Change-Id: Ie6729733ae3fe96e0a4542775dfe6c9266c9222a

5 years agoElementary : fix migration build errors
Yeongjong Lee [Wed, 27 Mar 2019 07:35:58 +0000 (16:35 +0900)]
Elementary : fix migration build errors

@tizen_fix

Change-Id: Iad251f98ec2706deb69262ef721d526b9fcbe4b1

5 years agomigration: remove beta tags.
Yeongjong Lee [Wed, 27 Mar 2019 07:33:56 +0000 (16:33 +0900)]
migration: remove beta tags.

It will fix build error

@tizen_fix

Change-Id: I87a3905bd0f81af5f6182f8f96f29a71679169d3

5 years agoeolian_csharp: hide struct native representation inside the managed struct
Vitor Sousa [Mon, 25 Mar 2019 20:14:32 +0000 (17:14 -0300)]
eolian_csharp: hide struct native representation inside the managed struct

Summary:
Rework the struct binding generator to declare the native struct nested inside
the managed one.
This way native structs are less likely to cause confusion; for example
with an IDE that supports automatic completion.

Get rid of struct conversion class methods in favor of using (the already
generated) implicit conversion operators.

Depends on D8469

Reviewers: segfaultxavi, lauromoura, felipealmeida

Reviewed By: lauromoura

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8470

5 years agoeolian_csharp: make struct generator use indentation level from context
Vitor Sousa [Fri, 22 Mar 2019 21:47:27 +0000 (18:47 -0300)]
eolian_csharp: make struct generator use indentation level from context

Summary:
Make the struct generator (the first one to) extract indentation information
from the context, an effort for a future overall consistency in white space
generation.

Depends on D8468

Reviewers: lauromoura, segfaultxavi, felipealmeida

Reviewed By: lauromoura

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8469

5 years agoeolian_csharp: add indentation context
Vitor Sousa [Fri, 22 Mar 2019 21:37:50 +0000 (18:37 -0300)]
eolian_csharp: add indentation context

Summary:
Also, use new context class for cleaner constructs.

Also, make functions receive context objects by reference to avoid
unnecessary object copies (since context objects are bigger now).

This commit contains preparation structures for a future overhaul of
white space generation.

Depends on D8467

Test Plan: ninja test

Reviewers: felipealmeida, lauromoura

Reviewed By: lauromoura

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8468

5 years agoeolian_cxx: rework context code, also more flexible scope indentation generator
Vitor Sousa [Fri, 22 Mar 2019 21:34:29 +0000 (18:34 -0300)]
eolian_cxx: rework context code, also more flexible scope indentation generator

Summary:
Context management classes now store constant copies of the tags they
are supposed to hold. So, they are immutable objects now.
Functions that manipulate context create new object copies as needed.
`constexpr` was added for class constructor and functions.

Indentation generator now used four space by default.
Also, the amount of desired spaces can now be specified at call site.

Added helper methods to return 'indented' and 'unindented' version of a
given indentation generator object.

Test Plan: ninja test

Reviewers: felipealmeida, lauromoura

Reviewed By: lauromoura

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8467

5 years agoefl-csharp: Remove spaces from type strings in generator.
Lauro Moura [Tue, 26 Mar 2019 17:45:18 +0000 (14:45 -0300)]
efl-csharp: Remove spaces from type strings in generator.

Summary:
Previously, the type_impl and marshall_type_impl generators relied on a
type mismatch in the match table to fallback to the else branch in the
match check to actually print the type string. This was achieved by
adding the " " prefix to the type.

This commit changes this behavior to invoke a proper visitor just to
print and makes both generators return trimmed type strings.

This will help conforming to the C# coding conventions.

Test Plan: run test suite

Reviewers: felipealmeida, vitor.sousa

Reviewed By: felipealmeida, vitor.sousa

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8441

5 years agoelementary: fix ATSPI reflection API warning.
Cedric BAIL [Tue, 26 Mar 2019 16:05:40 +0000 (12:05 -0400)]
elementary: fix ATSPI reflection API warning.

Summary: Depends on D8451

Reviewers: zmike, bu5hm4n, stefan_schmidt, devilhorns

Reviewed By: devilhorns

Subscribers: #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8452

5 years agoeina: always initialize i.
Cedric BAIL [Tue, 26 Mar 2019 16:05:02 +0000 (12:05 -0400)]
eina: always initialize i.

Reviewers: zmike, bu5hm4n, stefan_schmidt, devilhorns

Reviewed By: devilhorns

Subscribers: #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8451

5 years agoevil : remove mkstemp, which is already defined in mingw-w64
Vincent Torri [Tue, 26 Mar 2019 14:10:54 +0000 (10:10 -0400)]
evil : remove mkstemp, which is already defined in mingw-w64

Summary: mkstemp is already defined in mingw-w64

Test Plan: compilation

Reviewers: zmike

Reviewed By: zmike

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8454

5 years agoevas_3d: stop just segfaulting straight away
Marcel Hollerbach [Tue, 26 Mar 2019 14:09:42 +0000 (10:09 -0400)]
evas_3d: stop just segfaulting straight away

Summary:
there have been wrong function calls, that did not work at all, since
the function pointer had the wrong type. This fixes the segfaulting
examples of evas3d. However, they still do not render, at least, they
don't crash anymore.

Depends on D8381

Reviewers: cedric, segfaultxavi, zmike, stefan_schmidt

Reviewed By: zmike

Subscribers: #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8382

5 years agoefl_ui_tab_pager: fixed issue where tab_pager does not work
Bowon Ryu [Tue, 26 Mar 2019 05:20:27 +0000 (14:20 +0900)]
efl_ui_tab_pager: fixed issue where tab_pager does not work

Summary:
I fixed the tab_pager behavior problem caused by changing efl interface.
and I have also fixed some sample code errors.
 - using efl_file_load for efl_file_set
 - using efl_del for unpacked tab_page
 - disable unimplemented features

Test Plan: elementary_test -to efl.ui.tab_pager

Reviewers: Jaehyun_Cho

Reviewed By: Jaehyun_Cho

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8471

5 years agotests: add function to do "real" timers
Mike Blumenkrantz [Thu, 21 Mar 2019 17:24:47 +0000 (13:24 -0400)]
tests: add function to do "real" timers

in very specific cases it's necessary to match the exact timing of
internal functionality, so add a function to provide that capability

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8439

5 years agotests: make elm_win visible during tests
Mike Blumenkrantz [Thu, 21 Mar 2019 17:08:56 +0000 (13:08 -0400)]
tests: make elm_win visible during tests

elm_win has some internal locking to avoid doing sizing and visibility
changes until pre-render to save some calculations. this makes triggering
ui events on objects impossible, as they will not be visible.

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8438

5 years agomeson build : remove duplicate check of dirent.h
Vincent Torri [Mon, 25 Mar 2019 14:10:47 +0000 (10:10 -0400)]
meson build : remove duplicate check of dirent.h

Reviewers: zmike

Reviewed By: zmike

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8460

5 years agomeson build : remove duplicate check of strerror_r
Vincent Torri [Mon, 25 Mar 2019 14:10:37 +0000 (10:10 -0400)]
meson build : remove duplicate check of strerror_r

Summary: strerror_r is checked twice in headers_check

Reviewers: zmike

Reviewed By: zmike

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Differential Revision: https://phab.enlightenment.org/D8453

5 years agoelm_entry: fix a bug in moving the magnifier
WooHyun Jung [Mon, 25 Mar 2019 12:52:10 +0000 (08:52 -0400)]
elm_entry: fix a bug in moving the magnifier

Summary:
When moving the magnifier in elm_entry, _magnifier_move
function occured infinite loop problem by calling
evas_object_move twice with different positions.
So, I changed it to call evas_object_move once.

ref T7202

Test Plan:
1. Set profile to "mobile"
2. elementary_test
3. entry
4. long press mouse button on elm_entry
5. when the magnifier is shown, try to move it

Reviewers: zmike, bu5hm4n, bowonryu, id213sin

Reviewed By: zmike

Subscribers: cedric, #reviewers, #committers

Tags: #efl

Maniphest Tasks: T7202

Differential Revision: https://phab.enlightenment.org/D8462

5 years agoefl_ui_image: remove job in sizing calc.
Hermet Park [Mon, 25 Mar 2019 10:46:24 +0000 (19:46 +0900)]
efl_ui_image: remove job in sizing calc.

Originally, this job task was introduced for trick optimization to avoid
unnecessary duplicated compuation.

But this introduced a regression bug as well, image geometry updation was
delayed by this.

So, we remove the job here.

@fix T7360

Change-Id: I6a11744777186b486e81b87ea5293fa564c732ed

5 years agoefl_mono: remove unnecessary external function import
Jaehyun Cho [Mon, 25 Mar 2019 09:58:03 +0000 (18:58 +0900)]
efl_mono: remove unnecessary external function import

ecore_init, ecore_shutdown, elm_init, elm_policy_set, elm_shutdown,
elm_exit are imported in efl_all.cs.
Moreover, efl_csharp_application.cs is using Efl.UnsafeNativeMethods.
Therefore, the unnecessary external function import is removed in
efl_all.cs.

5 years agoefl_ui_container_layout: use correct property
Yeongjong Lee [Mon, 25 Mar 2019 09:56:41 +0000 (09:56 +0000)]
efl_ui_container_layout: use correct property

The `if` condition check whether horizontal, vertical aspect are greater than 0.
`aspect` is correct here.

this patch fixes T7756

Reviewed-by: Marcel Hollerbach <mail@marcel-hollerbach.de>
Differential Revision: https://phab.enlightenment.org/D8461

5 years agoelm_gengrid: speed up update order calls
Marcel Hollerbach [Fri, 15 Mar 2019 13:12:04 +0000 (14:12 +0100)]
elm_gengrid: speed up update order calls

update order can be quite expensive, so this here tries to skip it as
often as possible.

ref T7384

Reviewed-by: SangHyeon Jade Lee <sh10233.lee@samsung.com>
Differential Revision: https://phab.enlightenment.org/D8367

Change-Id: I750d4ac6ebbb03f43a6727786316dd313021a87c

5 years agoelm_gengrid: rework focus registration
Marcel Hollerbach [Fri, 15 Mar 2019 12:37:33 +0000 (13:37 +0100)]
elm_gengrid: rework focus registration

before just everything has been registered, now only the realized items
are registered.

ref T7384

Reviewed-by: Cedric BAIL <cedric.bail@free.fr>
Differential Revision: https://phab.enlightenment.org/D8366

Change-Id: I15fe4b231ba5660c4edffceaadd199d7c31ba96f