* give the possibility to set edje_cc (useful when cross compiling)
* clean .edj files
Next part: use eina_prefix to select edj and png files
SVN revision: 68130
I was obliged to do that because almost all the headers are
included in edje_private.h... That way to include headers is
fine on unix, but when doing also compilation on Windows,
it's just a nightmare
SVN revision: 68111
This make it possible to completly disable signal broadcasting as this
new behaviour broke Edje 1.0 file. It's also now possible to use the
same group in different part in the same parent group without any issue.
I am tempted to backport this patch to 1.1 branch as it would make it
play nicely with file coming from Edje 1.0.
Another issue that this patch fix is that I did increment the minor version
as we really have add a lot of addition since Edje 1.1 and Edje file build
with trunk may not play well anymore on Edje 1.1.
SVN revision: 67936
This patch try to prevent the broadcasting of targeted message. This should minimize
the problem generated on edje sub GROUP that didn't expect to see that much message
coming to them. It just a minimization of the problem, as message that don't target
explicitly a part are still propagated and can still break your edje usage from 1.0
to 1.1 version.
SVN revision: 67830
edje_object_text_insert_filter_callback_add, a filter function
should check the type(format, text, or markup) of text for correct
filtering. But in markup filter function(added by new API), text is
always markup so no need to check the type of text.
SVN revision: 67724
This should probably become the new standard for coords as it doesn't
clash with the damn y1 posix function.
Thasks to Gustavo for the naming.
SVN revision: 67325
Long time ago, in
http://www.mail-archive.com/enlightenment-devel@lists.sourceforge.net/msg32795.html
mail thread and IRC,
I talked with about problem of asynchronous event API such as
ecore_imf_context_commit_event_add,
ecore_imf_context_preedit_changed_event_add, so on.
In short, The problem is that key event and text_set APIs are processed
immediately, but commit event and preedit changed event is processed
asynchronously because those APIs add each event to ecore event queue.
To fix these problems, I've decided to create synchronous event APIs such
as ecore_imf_context_event_callback_add, del and call.
For considering compatibility, sync and async event callback functions are
used in xim and scim immodule.
SVN revision: 67290
be 1.1 (or 1.5) for the last release. too late. THIS is why i'm sick
and tired of all the bloody separate libs that have to be versiioned
and build and released separately. :( too many places to go fix up per
release.
SVN revision: 67284
Global shadow warnings are annoying and thus will be ignored at the moment,
There are other issues as well though, for example, not using cpp_token_file.
I don't know if that's intended or not, so I won't just suppress the warning.
SVN revision: 67242
allows us to multiple a minimim size explicitly for min size calc so
we can do things like have content slide open/closed properly.
SVN revision: 67197
This is so that future scripts will still work with old libraries,
and lets us add the "host can provide Lua API" feature soon.
Also some more comments.
SVN revision: 66961
NOTE: it's just a partial revert of previous patch by
raster. Without that, some text were flickering going
on and off for sometime. I didn't take the time to
understand why, but by forcing the recalc it permanently
solve the issue.
SVN revision: 66903
NOTE: I am still wondering what is the cost difference between
forcing a request to eet_open and calling stat. If someone has some
time to benchmark, feel free to do so and report on e-devel ml.
SVN revision: 66902
differently. When inputting Korean text, preediting text will be shown
as selected. When inputting Japaneses text, preediting text will be
shown with underline. (Sometimes shown as selected for changing whole
preediting text to another text)
SVN revision: 66580
r66257 <- another small fix on the real problem
r66250 <- a small fix on the real problem
r66242 <- the real problem
jaewhan - your commit yesterday (r66242) has made edje_cc very
unstable and it will randomly segv (sometimes yes, sometimes no). as
such it's at the point i can't even compile e and elm without it
segving somewhere during build, so this gets backed out. review your
change carefully and look for issues.
SVN revision: 66265
Lately, raster removed the code about the prohibition of type-change in
group inherit.
But about the "part" of different type, the data structure of the their
"description" is different.
So if the type is changed, it have to be reallocated. Current, it is not.
At first, we have to remove the lookups. If we don't, when lookup module
executes, the memory
may be broken. So I removed all lookups for reallocated description before
it is reallocated.
And I changed all description of the "part" is reallocated when the type is
changed.
The attribute of the "part" is remained. Just it reallocated the part of
**_Spec_**.
SVN revision: 66242
NOTE: This was necessary for solving issue with the new CURRENT feature. I
don't like this massive change, but there is no way around. This patch is
only the first step, I will wait the full night before completly fixing
the issue with CURRENT.
WARNING: If this patch doesn't break svn, you must feel lucky and go play
money games. In all other case, please report any issue to the developper
mailing-list.
SVN revision: 65619
and.... u get the idea. this made an n^m list of messages... where n
was 3 of messages sent and m was # of child parts (42 of them)... this
caused a silly 3 of timers to be allocated... don't ask how many. in a
simple snapshot i saw 101mb of timers allocated... and i was just
starting... anyway - this makes the propagatiopn not propagate down
and then back up again... and it only needs 1 timer allocated to
handle a re-schedule of processing messages. not N. "leak" that was
just a massive memory spike) is now fixed.
SVN revision: 65571
AND signals should propagate to children the SAME WAY - that means
children you added explicitly in the edje file and not form c code get
propagated to.
SVN revision: 65384
The very original lua code from waaaay back when,
would create the various timer objects as members of the evas class,
which would mean you could call evas functions on the timer objects.
Not good.
Now they have their own classes.
However, you still have to create them by calling functions in the
edje class, there should probably be an ecore class for that instead.
SVN revision: 65354
Actually, if rel1.to_x is not used in the matched description of the parent group, rel1.to_x: ""; doesn't need to be used in the description of the child group.
Because its state is already default. But sometime we may want to use it explicitly. So I add this code in order to cover that case.
SVN revision: 65160
the edje group is swallowed in that, and the bubbles lua group
is swallowed into the edje group.
Demonstrates all ways of lua swallowing, except direct lua to lua.
SVN revision: 65140
This happens during startup at least and did cause problems.
Restrict image loading to images from the same edje file we are in, as per rasters request.
Some minor cleanups.
This commit fixes the last of the known problems.
So it's ready for release, except for adding docs.
I'll be adding docs over the next week anyway.
SVN revision: 65136
I don't know doxygen, and doxygen don't seem to quite understand what I want.
Which is to document the lua side of things, but having the docs near the C
functions that define them. Help me out please doxygen experts.
Note the problem with edje:rect().
Or I could just write one big blob. Ewwww.
SVN revision: 65097
and returning a table of results.
This replaces a whole host of specialised functions, and gives some flexibility.
As promised, I did not forget about the const castings, they went away as expected.
SVN revision: 65084
had to add eina_init() and shutdown and init refcounting and log
domain creating and add a macro.. so i can finally use it? eina being
included here actually was never needed... but since include was there
i could use it...
just saying - using eina_log is not convenient. it doesn't save time.
it's benefits are dubious (knowing what envv ar to set to what values
to get log output is black magic as u need to know specific log level
values and log domain names which u only find out by digging thru
code). my point -> if u want eina_log used more.. it needs to be AS
EASY as the printf. :)
SVN revision: 64937
Also fix up some warnings pointed out by Mike Blumenkrantz.
I'm using the "I committed this from my USA server to sneak it past the deadline" excuse. :-P
SVN revision: 64857
From: Govindaraju S M <govi.sm@samsung.com>
Subject: edje multisense patch for sound sample, tone and haptic play
... This is the beginning of sound (and haptic) support in Edje - it
works, but only at certain basic sample playback levels. more will come.
SVN revision: 64731
Currently, when the part in edc move by transition, it moves from the
one of the states to another of the states even if it is ainimating.
We need it to move from current position, because the animation is not
natural and smooth. So I made the "CURRENT" option.
SVN revision: 64675
Currently, when the part in edc move by transition, it moves from the
one of the states to another of the states even if it is ainimating.
We need it to move from current position, because the animation is not
natural and smooth. So I made the "CURRENT" option.
And I removed the unnecessary tab for indentation.
SVN revision: 64545
elm_entry_autocapital_type_set API is high level API to call ecore_imf_context_autocapital_type_set.
Application programmer can choose the type of autocapitalization such as
WORD, SENTENCE, ALLCHARACTER through this API.
elm_entry_input_panel_enabled_set API is for setting whether input panel
(virtual keyboard) should be appeared when entry has a focus or pressed.
It can be used by dialer or calculator application programmer because those
application prefer to use its keypad NOT virtual keyboard supported by system.
They want to use entry to show the cursor for providing the cursor handling
method but don't want to appear system keyboard.
SVN revision: 64423
Subject: [E-devel] [Patch] change rel1.to to default when using group
inherit
When using group inherit, if rel1.to is set in parent group, we cannot change the part name
of rel1.to to default (the default is that dont have the part name) in child group.
So I changed edje_cc_out.c. After patching, we can change rel1.to to
default by using "". (ex > rel1.to: "";)
SVN revision: 64171
2011/9/30 Jaehwan Kim <jae.hwan.kim@samsung.com>:
>
> I found some bugs in edje_cc_handlers.c.
>
> One is about data.item. It is saved in hash list.
> When it is saved, I freed the memory of the key.
> But the memory is used. so I deleted the free(key).
>
> Second is it is possible in
st_collections_group_parts_part_description_state.
> So I changed the location of the conditional statement.
>
> Please apply this patch :)
>
> Jaehwan Kim.
SVN revision: 63822
Info added includes: cursor position of the action, content related to the
action, and the action that happened (added/removed).
Also fixed tab insertion to delete the current selection.
SVN revision: 63773
Subject: [E-devel] [PATCH] update cursor location when the position of
entry is changed
The cursor location information should be transferred when the position of
entry is changed.
For example, an entry can be moved by comformant, rotate event.
The transferred cursor location is used for letting candidate word window
move the appropriate position.
SVN revision: 63660
Jihoon Kim
WooHyun Jung
Vincent Torri
Tom Hacohen
David Walter Seikel
Cedric BAIL
Carsten Haitzler
Mike Blumenkrantz
Davide Andreoli
Iván Briano
Sebastian Dransfeld
Youness Alaoui
Daniel Juyung Seo
Gustavo Lima Chaves
Jaehwan Kim
Prince Kumar Dubey
Bruno Dilly
Jaehwan Kim
Mike McCormack
Boris Faure
Jihoon Kim