Tom Hacohen [Mon, 4 Aug 2014 12:29:20 +0000 (13:29 +0100)]
Eolian CXX: Get rid of built sources for the callbacks test sources.
This is not needed now that dependencies are well in place.
Thiep Ha [Mon, 4 Aug 2014 10:01:51 +0000 (11:01 +0100)]
Evas textblock: Correct word start/end moving at new line or line begins with spaces
Summary:
Word start/end works incorrectly when it goes to new line or line begins with spaces.
Ex: In elementary_test/Entry, place cursor at the end of line, press ctrl + right arrow keys: cursor moves to begin of next line. In this case, cursor should move to end of 1st word in next line.
Ex2: In elementary_test/Entry, add some spaces to begin of 2nd line (" uses markup"), place cursor at the first word ("uses"), press ctrl + left arrow keys twice, cursor moves to begin of 2nd line. In this case, cursor should move to begin of last word in 1st line.
This patch provides a fix by considerring next/previous text node to move cursor to correct place.
@fix
Reviewers: woohyun, raster, tasn
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D1140
Daniel Kolesa [Mon, 4 Aug 2014 09:56:15 +0000 (10:56 +0100)]
eolian: fread fix
Instead of 1 element of size N, we should read N elements of size 1 in
order to sensibly compare the result with the result of calling ftell.
Cedric BAIL [Mon, 4 Aug 2014 09:55:05 +0000 (11:55 +0200)]
ecore_x: don't go below 0.
It seems that when things go wrong it does happen that we start with i == 0. It
will then have been walking into info[-1] 'happily'. Changing the test, to first
decrement and then compare should stop that issue.
This should @fix T1467.
Daniel Kolesa [Mon, 4 Aug 2014 09:44:35 +0000 (10:44 +0100)]
ethumbd_slave: fix coverity defect
Fixes CID 1193232 (string not null terminated)
Jean Guyomarc'h [Mon, 4 Aug 2014 09:34:51 +0000 (11:34 +0200)]
doxygen: sanitize Eo Tutorial
Summary: The Eo tutorial had a tough readability. It is better now: stars are not messing around with the code, plus code does not collide anymore with a bullet in a bullet list.
Reviewers: cedric, JackDanielZ, tasn
Reviewed By: tasn
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D1230
Daniel Kolesa [Mon, 4 Aug 2014 09:26:06 +0000 (10:26 +0100)]
edje_cc_handlers: fix coverity defects
Fixes CID 1193224, 1193225, 1193226, 1193227, 1193230 (resource leak)
Daniel Kolesa [Mon, 4 Aug 2014 08:58:32 +0000 (09:58 +0100)]
eolian/generator: coverity fix
Fixes CID 1229130 (ignoring number of bytes read)
bluezery [Sat, 2 Aug 2014 19:09:57 +0000 (21:09 +0200)]
Eina: Change from "" to <> for header inclusion
Summary:
All eina_xxx.h are expored headers.
Using system path to search eina headers seems to be better.
Reviewers: raster, stefan_schmidt, stefan, cedric
Subscribers: stefan_schmidt, cedric
Differential Revision: https://phab.enlightenment.org/D1079
Carsten Haitzler (Rasterman) [Sat, 2 Aug 2014 13:54:23 +0000 (22:54 +0900)]
fix warning about unused var
Carsten Haitzler (Rasterman) [Sat, 2 Aug 2014 04:16:27 +0000 (13:16 +0900)]
ecore-evas - fix continual wakeups with unchanged canvases
this fixes a nasty bug where ecore-evas forces mainloop spins all the
time due to trying to align rendraws to animator boundaries. this
requires an extra evas feature to work that i just put in.
@fix
Carsten Haitzler (Rasterman) [Sat, 2 Aug 2014 04:15:59 +0000 (13:15 +0900)]
add evas changed get in order to help fix ecore-evas bug
this does add an api, but it is to fix another bug in ecore-evas
Carsten Haitzler (Rasterman) [Sat, 2 Aug 2014 04:15:14 +0000 (13:15 +0900)]
ecore-x vsync tool. fix multi-client handling so ticking continues
a mistaking in using the wrong tick var caused ticking to stop when a
client that is ticking exits while another ticks. this fixes it.
Carsten Haitzler (Rasterman) [Sat, 2 Aug 2014 03:54:34 +0000 (12:54 +0900)]
clean vsync code - use do+while not goto.
Carsten Haitzler (Rasterman) [Sat, 2 Aug 2014 03:03:35 +0000 (12:03 +0900)]
ecore-x-vsync - rearragnge to move vsync event input to thread
this uses a thread to collect vsync input events and filter them
before forwarding them to the mainloop (as a double timestamp). this
means wakeups only happen for the actual vsync and thus animator and
not for other screens we are filtering out anyway. this should fix the
continual animator wakeups that happen if you have a dri/drm based
driver and > 1 screen.
Stefan Schmidt [Fri, 1 Aug 2014 21:59:48 +0000 (23:59 +0200)]
build: Bring coverage back to a working state
With commit
6030b9de79a3403d42a01432b67e0fa7a3948149 the internal name EFL
name was changed to EFLALL but the needed CFLAGS and LDFLAGS for coverage
have not been adjusted. Thus it was simply no longer producing the gcda
files needed by lcov.
All back now and it shows an amazing jump in our coverage to:
Overall coverage rate:
lines......: 31.6% (45827 of 144975 lines)
functions..: 41.1% (5620 of 13684 functions)
MinJeong Kim [Fri, 1 Aug 2014 12:25:52 +0000 (14:25 +0200)]
build/drm: Fix missing ecore-drm for evas_drm engine
Summary:
With removing of pkgconfig checking on EVAS_CHECK_ENGINE for drm,
evas_drm engine need to setup libs including internal ecore-drm.
But, the evas_drm engine have missed ecore-drm libs because it have been
done after finishing setup library of evas.
This revision moves setup dependendency for ecore-drm into proper place.
Test Plan:
1. Build EFL with --enable-drm
2. ELM_ENGINE=drm E_WL_FORCE=wayland_shm enlightement_start
Reviewers: gwanglim, devilhorns, stefan_schmidt
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D1249
Carsten Haitzler (Rasterman) [Fri, 1 Aug 2014 09:59:18 +0000 (18:59 +0900)]
fix wakeups to not happen for now until i fix vsync
Stefan Schmidt [Fri, 1 Aug 2014 08:21:15 +0000 (10:21 +0200)]
release: Update NEWS and bump version for 1.11.0-alpha1 release
Carsten Haitzler (Rasterman) [Fri, 1 Aug 2014 00:33:16 +0000 (09:33 +0900)]
ecore-x vsync nvidia glx tool - handle clients disabling first then enabling
Carsten Haitzler (Rasterman) [Fri, 1 Aug 2014 00:31:18 +0000 (09:31 +0900)]
ecore animator - this fixes suspended animators still waking up
@fix
this fixes a long standing issue where a suspended animator still is
waking up as originally suspended animators were expected to hang out
for a small time. as e's comp uses a suspended animator, this is a
problem as it causes continual wakupes every single frame (60hz or so)
with this suspended animator. this fixes that and accounts for
suspended animators with tick begin/end
Carsten Haitzler (Rasterman) [Thu, 31 Jul 2014 23:47:06 +0000 (08:47 +0900)]
ecore_x - vsync nvidia/glx service fix so it connects 1st not 2nd time
Carsten Haitzler (Rasterman) [Thu, 31 Jul 2014 14:16:40 +0000 (23:16 +0900)]
ecore_x_vsync - protect against silly clients asking for vsync lots
this allows a client to ask multiple times and have to release the
same number of times for vsync events to stop. covers possible misuse
Daniel Kolesa [Thu, 31 Jul 2014 13:26:52 +0000 (14:26 +0100)]
autotools: fix distcheck by removing Makevars in proper stage
Cedric BAIL [Thu, 31 Jul 2014 01:38:42 +0000 (03:38 +0200)]
autotools: po/Makevars is generated by configure, not Makefile.
If you did do a make clean, then you couldn't do a make afterward. Reason
was that po/Makevars is generated by configure and not Makefile. So it can
only be cleaned with make distclean and not make clean.
Stefan Schmidt [Thu, 31 Jul 2014 09:21:20 +0000 (11:21 +0200)]
tests/eolian_cxx: Package callback.eo file into tarball in all cases.
When running make distcheck with release mode enabled the callback.eo
file did not find its way into the tarball and failed the build.
Thiep Ha [Thu, 31 Jul 2014 03:28:42 +0000 (05:28 +0200)]
ecore_evas_extn: fix wrong plug image display after connect
Summary:
Plug image is displayed incorrect after connect to socket.
Test case: Run ecore_evas_extn_socket_example -> run ecore_evas_extn_plug_example -> click Change bg to change bg color.
Run 2nd ecore_evas_extn_plug_example. The plug area image of 2nd plug is incorrect display (different with 1st plug image).
Reason: When a plug connects to socket, socket sends incorrect buffer information.
Fix: Change buffer information.
@fix
Reviewers: Hermet, huchi
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D1232
Savio Sena [Thu, 31 Jul 2014 02:04:48 +0000 (23:04 -0300)]
autotools: Fix eolian_cxx tests.
Added callback.eo.c and callback.eo.h as dependencies to
eolian_cxx_test_callback.c and also added all generated files to CLEANFILES.
Carsten Haitzler (Rasterman) [Thu, 31 Jul 2014 01:47:05 +0000 (10:47 +0900)]
fix build dep problem with ecore-x
Carsten Haitzler (Rasterman) [Wed, 30 Jul 2014 12:33:25 +0000 (21:33 +0900)]
Revert "tests/eolian_cxx: Add another generated file to BUILT_SOURCES"
This reverts commit
577a6bfbba0a1cf19f394c66addf27b0332500d6.
This brings efl back to building - it entirely stopped building for me
with this change so even if this tried to fix something... it made
just regular building of efl at all cease to work here which i'd say
is bad (if this just fixes a dist problem - at least dist is broken,
but regular build is not - we are in better shape).
Carsten Haitzler (Rasterman) [Wed, 30 Jul 2014 07:16:45 +0000 (16:16 +0900)]
evas - gl-x11 - extensions - be more specific about video_sync extn
this should avoid false-positive detection for this extension - other
extensions share string content
Daniel Kolesa [Wed, 30 Jul 2014 12:33:52 +0000 (13:33 +0100)]
eolian: fix: remove unused function
Stefan Schmidt [Wed, 30 Jul 2014 08:53:15 +0000 (10:53 +0200)]
tests/eolian_cxx: Add another generated file to BUILT_SOURCES
If not put into BUILT_SOURCES it will not be added to CLEANFILES
and thus staying around as uncleaned file which make distcheck
complain about it.
Stefan Schmidt [Wed, 30 Jul 2014 07:19:53 +0000 (09:19 +0200)]
build: Clean Makevar now that we generate one on our own
Since we generate our own Makevar for the po subdir we also need to clean
it up. If not distcheck complains about uncleaned files.
Cedric BAIL [Wed, 30 Jul 2014 01:47:31 +0000 (03:47 +0200)]
ecore_evas_extn: do not double close fd in case of memory starvation.
In case eina_stringshare_add was failing it could lead to a double close
on lockfd. This was reported by coverity: CID1039885
@fix
Cedric BAIL [Wed, 30 Jul 2014 01:41:46 +0000 (03:41 +0200)]
autotools: shutup stupid warning.
Tom Hacohen [Wed, 30 Jul 2014 01:21:18 +0000 (02:21 +0100)]
Eldbus C++: Don't generate C++ bindings when C++ is disabled.
Carsten Haitzler (Rasterman) [Tue, 29 Jul 2014 23:38:21 +0000 (08:38 +0900)]
fix vsync display env var handling to be right
Chris Michael [Tue, 29 Jul 2014 21:02:04 +0000 (17:02 -0400)]
evas-wayland-shm: Free allocated rectangle after sending damages
We allocate a new eina_rectangle here, but we never free it after
sending damages to the surface.
@fix
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Stefan Schmidt [Tue, 29 Jul 2014 19:09:16 +0000 (21:09 +0200)]
build: Make sure xgettext_wrapper.sh gets packaged into the tarball
We need this in the tarball when not running from git. Also fixes
make distcheck.
Stefan Schmidt [Tue, 29 Jul 2014 15:05:30 +0000 (17:05 +0200)]
evas/engine_gl_generic: Make sure we epend on the needed header for compiling.
Fixup for
a3a15b1986392bcc48affbab5c502013a83bfddd which removed the header
from installation. We need to have it in SOURCES now during distcheck.
Stefan Schmidt [Tue, 29 Jul 2014 14:30:37 +0000 (16:30 +0200)]
Revert "configure.ac: generate Makevars in the correct directory even with out-of-tree build"
This reverts commit
dd37d2bc07f839cd3c74b02e713ffc6534af71b2.
This breaks make distcheck. Looking at this commit I really wonder if it does
anything good. It seems to work for po_makefile_in. It also breaks for
distcheck which is using out of tree builds in the _build folder.
If someone can explain me why something like this is needed for makevars I
want to hear it. getting a fix in that does not break distcheck would be fine
I guess.
Carsten Haitzler (Rasterman) [Tue, 29 Jul 2014 14:22:06 +0000 (23:22 +0900)]
ecore-x-vsync - oops - fixed, but debug paths left in - remove..
Carsten Haitzler (Rasterman) [Tue, 29 Jul 2014 14:19:52 +0000 (23:19 +0900)]
ecore-x-vsync - found issue when vt switchign with nvidia - workaround
Carsten Haitzler (Rasterman) [Tue, 29 Jul 2014 12:34:39 +0000 (21:34 +0900)]
fix parallel build for ecore_x_vsync tool
Carsten Haitzler (Rasterman) [Tue, 29 Jul 2014 11:00:53 +0000 (20:00 +0900)]
ecore_x - vsync fix drm support to work again
due to mesa changes to hide dri2 symbols, i have had to work on a fix
that makes this work again by going right to drm. now it works and
animators shoudl be vsynced on drm drivers if possible (only 1 card -
use card 0). already existing nvidia solution that uses a lot more
memory is there. others - no support. timers only
Tom Hacohen [Tue, 29 Jul 2014 09:06:21 +0000 (10:06 +0100)]
Evas GL Generic: Don't install an internal header.
I'm not 100% sure this header is internal, but it depends on internal
headers, so it should either be internal or massively re-worked.
Anyhow, it has no business being installed until the includes are
fixed.
Daniel Kolesa [Tue, 29 Jul 2014 09:05:06 +0000 (10:05 +0100)]
eolian: remove irrelevant comments
This functionality was already done previously.
perepelits.m [Tue, 29 Jul 2014 07:37:25 +0000 (16:37 +0900)]
[Evas\evas-3d\example] Add example of 3d-shooter
Reviewers: cedric, raster, Hermet
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D1215
Wonguk Jeong [Tue, 29 Jul 2014 00:11:17 +0000 (02:11 +0200)]
ecore: fix example build fail (ecore_compose_get_example)
Carsten Haitzler (Rasterman) [Sun, 27 Jul 2014 05:22:27 +0000 (14:22 +0900)]
ecore x vsync animator support - add a glx based vsync ticker
this adds a slave process that is useful on nvidia drivers as there
isn't another way to get vsync evenys (that i know about). i need to
make another slave process to that includes a dri2 protocol
implementation since mesa has now hidden its dri2 symbols.
Mike Blumenkrantz [Mon, 28 Jul 2014 13:47:48 +0000 (09:47 -0400)]
add simple ecore_con_url test suite
Mike Blumenkrantz [Mon, 28 Jul 2014 13:45:52 +0000 (09:45 -0400)]
call ecore_init() for ecore_con_url_init()
ecore functions are used by this api, so ecore must be initialized
Mike Blumenkrantz [Mon, 28 Jul 2014 13:14:54 +0000 (09:14 -0400)]
remove unnecessary header include from ecore-con test
Daniel Kolesa [Mon, 28 Jul 2014 10:57:37 +0000 (11:57 +0100)]
eolian: fix: remove leftover keywords
Daniel Kolesa [Mon, 28 Jul 2014 09:16:15 +0000 (10:16 +0100)]
eolian: preliminary eo file support for builtin true/false/null
Real API will use the new expression system, but that won't get into 1.11.
Jihoon Kim [Mon, 28 Jul 2014 07:01:07 +0000 (16:01 +0900)]
ecore_imf: documentation about ECORE_IMF_INPUT_PANEL_LAYOUT_NUMBERONLY enums
Mike Blumenkrantz [Sun, 27 Jul 2014 22:51:42 +0000 (18:51 -0400)]
Revert "eina: Eina_Module - check whether the file exists or not, before memory allocation of Eina_Module"
This reverts commit
971589bf56b7b51eba305dfb5e544e6436740534.
this completely breaks existing eina_module usage. elm_test->image remote
Carsten Haitzler (Rasterman) [Sat, 26 Jul 2014 00:19:30 +0000 (09:19 +0900)]
move ecore documentation not in headers to .h files for consistency
Carsten Haitzler (Rasterman) [Sat, 26 Jul 2014 00:15:47 +0000 (09:15 +0900)]
add function to set ecore loop time
@feature
this allows you to set the ecore loop time. only useful in trying to
get hyper-accurate frame timings from sources when doin a custom tick
source.
Carsten Haitzler (Rasterman) [Sat, 26 Jul 2014 00:02:56 +0000 (09:02 +0900)]
ecore animator - adjust ecore loop time when animtor ticks to be exact
this adjusts ecore loop time very slightlye tp be the "Exact" timepoint
when then animator timer, if timer is used, should have gone off. this
should make animations more precise.
@feature
Cedric BAIL [Fri, 25 Jul 2014 18:23:42 +0000 (20:23 +0200)]
evas: move to SDL2 and use GL_Generic infrastructure.
Cedric BAIL [Fri, 25 Jul 2014 18:05:37 +0000 (20:05 +0200)]
evas: provide GetProcAddress by the engine.
Daniel Kolesa [Fri, 25 Jul 2014 16:11:20 +0000 (17:11 +0100)]
elua: make generator work again
Also adds to_array API to Lua eina iterators.
Daniel Kolesa [Fri, 25 Jul 2014 15:54:32 +0000 (16:54 +0100)]
elua: update eolian bindings
Thiep Ha [Fri, 25 Jul 2014 15:06:37 +0000 (16:06 +0100)]
Evas textblock: Fix memory leak
Summary:
In evas_textblock_cursor_word_end function, the breaks' memory is allocated but not freed when cursor position is equal to text length.
Fix: Remove memory allocating.
@fix
Reviewers: raster, tasn
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D1225
Savio Sena [Fri, 25 Jul 2014 14:32:02 +0000 (11:32 -0300)]
eolian-cxx: Added generation of call_constructor for parent_type.
Daniel Kolesa [Fri, 25 Jul 2014 13:51:13 +0000 (14:51 +0100)]
eolian: fix type parsing
I reorganized what's allowed and what is not a bit, particularly, strictened the
rules for named structs and loosened the rules for anonymous structs, and refined
them for void types. It should be all correct now.
Daniel Kolesa [Fri, 25 Jul 2014 13:31:39 +0000 (14:31 +0100)]
eolian: doc typo fix
Daniel Kolesa [Fri, 25 Jul 2014 13:25:45 +0000 (14:25 +0100)]
eolian: clarify documentation
Jean Guyomarc'h [Fri, 25 Jul 2014 12:58:03 +0000 (13:58 +0100)]
Fix Elua compiling on OSX
Reviewers: cedric, raster, raoulh, q66
Reviewed By: q66
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D1231
Jean Guyomarc'h [Fri, 25 Jul 2014 10:08:47 +0000 (11:08 +0100)]
Unregister eina_log domain
Reviewers: cedric, q66
Reviewed By: q66
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D1233
Carsten Haitzler (Rasterman) [Fri, 25 Jul 2014 05:45:50 +0000 (14:45 +0900)]
ecore-evas - enable attempting to turn on dri2 vsync
this enables the x11 dri2 vsync enabling code for x11 ecore-evas
Savio Sena [Thu, 24 Jul 2014 22:47:27 +0000 (19:47 -0300)]
eolian-cxx: Removed a warning from the example.
Savio Sena [Thu, 24 Jul 2014 22:35:12 +0000 (19:35 -0300)]
autotools: Fixed 'make dist' for Eolian-Cxx
Added .eo to EXTRA_DIST and *_stub.h to example's source lists.
Savio Sena [Thu, 24 Jul 2014 19:47:35 +0000 (16:47 -0300)]
eolian-cxx: Added example of Eo implementation in C++.
Instead of implementing Eo functions in C we do it in C++ simply.
Savio Sena [Thu, 24 Jul 2014 19:03:02 +0000 (16:03 -0300)]
eolian-cxx: Fixed inheritance operations generation rules.
The generated code was not converting the types from C to C++ correctly.
Savio Sena [Thu, 24 Jul 2014 19:02:09 +0000 (16:02 -0300)]
eo-cxx: Added _release to efl::eo::inherit.
Just like previously added in efl::eo::base.
Savio Sena [Thu, 24 Jul 2014 19:00:10 +0000 (16:00 -0300)]
eolian-cxx: Added grammar rule parameters_cxx_list.
Now with the correct semantics: list parameters converting to C++.
Savio Sena [Thu, 24 Jul 2014 18:54:04 +0000 (15:54 -0300)]
eo-cxx: Added generic parent_type and parent_set.
This allows us to use any type implementing _eo_ptr() with the parent
idiom "efl::eo::parent = obj" and also with efl::eo::inherit parent_set.
Savio Sena [Thu, 24 Jul 2014 18:49:54 +0000 (15:49 -0300)]
eolian-cxx: Renamed parameters_cxx_list -> constructor_parameters_list
Savio Sena [Thu, 24 Jul 2014 18:45:53 +0000 (15:45 -0300)]
eolian-cxx: Removed unused code.
Sanitized grammar removing old code which is not unecessary for good.
Savio Sena [Wed, 23 Jul 2014 20:07:40 +0000 (17:07 -0300)]
eolian-cxx: Fixed a bug in the generator.
It was not considering the variable name correctly.
Felipe Magno de Almeida [Thu, 24 Jul 2014 21:53:20 +0000 (18:53 -0300)]
eolian-cxx: Enhance test for callback use
Added test for use of Eolian callbacks in C++. Testing adding
callbacks, calling callbacks and deleting them.
Chris Michael [Thu, 24 Jul 2014 15:33:21 +0000 (11:33 -0400)]
ecore-drm: Fix improper shutdown for error cases in ecore_drm_init
If an error occurs during ecore_drm_init, we need to cleanup things in
the proper order. This fixes the case where ecore_event_shutdown and
ecore_shutdown May Not have been called (in previous revision) due to
the error conditions being out of order.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
Tom Hacohen [Thu, 24 Jul 2014 14:01:36 +0000 (15:01 +0100)]
Eo base: Fix marking of eolian functions with EOLIAN.
Tom Hacohen [Thu, 24 Jul 2014 13:41:23 +0000 (14:41 +0100)]
Eo base: Mark class methods.
Use the new @class attribute to mark class methods.
Daniel Kolesa [Thu, 24 Jul 2014 14:32:34 +0000 (15:32 +0100)]
eolian/generator: properly generate the data arg on class methods
For class methods, we need to generate unused void*, not a pointer to data struct.
Daniel Kolesa [Thu, 24 Jul 2014 13:01:33 +0000 (14:01 +0100)]
eolian: add test for class methods/properties
This also fixes a bug in the Eolian parser that handles class properties.
Daniel Kolesa [Thu, 24 Jul 2014 12:01:47 +0000 (13:01 +0100)]
eolian: parser support for constructor qualifier
This will be needed when the constructors {} section goes away.
Daniel Kolesa [Thu, 24 Jul 2014 11:58:29 +0000 (12:58 +0100)]
eolian: protected ctors
Daniel Kolesa [Thu, 24 Jul 2014 11:15:55 +0000 (12:15 +0100)]
eolian/generator: add support for class methods/properties
Daniel Kolesa [Thu, 24 Jul 2014 11:05:12 +0000 (12:05 +0100)]
eolian: add support for class functions
Daniel Zaoui [Thu, 24 Jul 2014 08:55:08 +0000 (11:55 +0300)]
Eolian: Generate typedefs and structs.
This patch permits the generation of typedefs and structs in the
legacy and Eo header files. It includes tests.
Daniel Zaoui [Mon, 21 Jul 2014 11:43:17 +0000 (14:43 +0300)]
Eolian: little refactoring of the C generator.
This change is a preparation for the generation of the types and a
little cleaning.
Carsten Haitzler (Rasterman) [Thu, 24 Jul 2014 08:18:35 +0000 (17:18 +0900)]
evas - ico loader - detect probable p0ng entry in ico file
ico files were defined to have bmp's in each key - in fact a subset of
them. unbenknownst to yours truly, vista now allows them to also be
pngs and thus the ico loader rejects them as corrupt. at least detect
it and complain right now
Carsten Haitzler (Rasterman) [Thu, 24 Jul 2014 06:22:40 +0000 (15:22 +0900)]
fix some formatting in ico loader
Jaehwan Kim [Thu, 24 Jul 2014 05:50:25 +0000 (14:50 +0900)]
edje: Add dpi feature.
Summary:
This dpi is used to get the scale for each collection.
If each collection has a described dpi, it calculates a proper scale
based on the dpi and dpi which is described in the collection.
@feature
Test Plan:
If add dpi to collection of edc, the edje will save the value as the dpi of the collection.
For example, if the dpi of your device is 100, you just set dpi: 100 in the collection of edc.
If the edj is loaded in another device(dpi is 200), it will scaled 2 times.
It is possible that the described dpi of application and theme are different.
In that case, application and theme have a different scale.
It makes the edj that made in different environment works in one device.
Reviewers: seoz, zmike, JackDanielZ, Hermet, woohyun, cedric, raster
Reviewed By: raster
Subscribers: cedric
Differential Revision: https://phab.enlightenment.org/D1190
Daniel Kolesa [Wed, 23 Jul 2014 23:32:06 +0000 (00:32 +0100)]
eolian/generator: fix coverity issues
Fixes CID 1228663 and 1228634.
Daniel Kolesa [Wed, 23 Jul 2014 18:38:08 +0000 (19:38 +0100)]
Revert "eina: new API: eina_file_path_basename"
This reverts commit
68282f8c42c6e0c3d66cf298e1a763a0d12b81cf.
This is actually not needed because of presence of basename in Evil.