platform/upstream/gstreamer.git
12 years agopo: Update .po files
Philippe Normand [Fri, 20 Jul 2012 07:38:47 +0000 (09:38 +0200)]
po: Update .po files

12 years agotests: gstbuffer: add tests for some mulitple map combinations
Mark Nauwelaerts [Thu, 19 Jul 2012 11:51:23 +0000 (13:51 +0200)]
tests: gstbuffer: add tests for some mulitple map combinations

12 years agominiobject: fix sharedness check
Wim Taymans [Thu, 19 Jul 2012 11:35:34 +0000 (13:35 +0200)]
miniobject: fix sharedness check

12 years agominiobject: refuse write when object is shared
Wim Taymans [Thu, 19 Jul 2012 11:20:27 +0000 (13:20 +0200)]
miniobject: refuse write when object is shared

In all cases, refuse to write an object when it is shared by more than one
object (also when the object was locked before).

See https://bugzilla.gnome.org/show_bug.cgi?id=679145

12 years agotests: gstbuffer: extend buffer copy test
Mark Nauwelaerts [Wed, 18 Jul 2012 13:21:33 +0000 (15:21 +0200)]
tests: gstbuffer: extend buffer copy test

12 years agoqueue2: set buffering-left to 0 on 100% buffering
Wim Taymans [Thu, 19 Jul 2012 10:42:05 +0000 (12:42 +0200)]
queue2: set buffering-left to 0 on 100% buffering

Set the buffering-left field in the query to 0 when we are completely buffered.
Improve the debug.

12 years agoqueue2: fix buffering query
Wim Taymans [Thu, 19 Jul 2012 10:14:29 +0000 (12:14 +0200)]
queue2: fix buffering query

Fix the buffering query, fill in the right buffering-left and estimated-total
values.

12 years agoqueue2: fix the buffering-left in the buffering message
Wim Taymans [Thu, 19 Jul 2012 08:54:16 +0000 (10:54 +0200)]
queue2: fix the buffering-left in the buffering message

The buffering-left field in the buffering message should contain a time estimate
in milliseconds about for long the buffering is going to take. We can calculate
this value when we do rate_estimates.

12 years agomessage: improve buffering message defaults
Wim Taymans [Thu, 19 Jul 2012 08:14:18 +0000 (10:14 +0200)]
message: improve buffering message defaults

Remove the estimated-total field, this should not be part of the buffering
message.
Set the default value of buffering-left to 0 when the percent is 100.

12 years agopad: fix debug line
Wim Taymans [Wed, 18 Jul 2012 15:44:54 +0000 (17:44 +0200)]
pad: fix debug line

Use QUERY_TYPE on query types.

12 years agoghostpad: remove custom function
Wim Taymans [Wed, 18 Jul 2012 15:35:58 +0000 (17:35 +0200)]
ghostpad: remove custom function

Remove custom pad functions, the default ones are better.

12 years agopad: add PROXY_SCHEDULING flag
Wim Taymans [Wed, 18 Jul 2012 15:30:30 +0000 (17:30 +0200)]
pad: add PROXY_SCHEDULING flag

Add a flag that makes the default query handler forward the scheduling query.

12 years agoutils: fix docs
Wim Taymans [Wed, 18 Jul 2012 15:30:04 +0000 (17:30 +0200)]
utils: fix docs

12 years agopad: improve query caps function
Wim Taymans [Wed, 18 Jul 2012 14:20:41 +0000 (16:20 +0200)]
pad: improve query caps function

In the proxy_query_caps function, also filter against the filter in the query.
We don't need to filter against the filter in the query anymore in the default
caps query function because we already did this in the proxy_query_caps.

12 years agobasesink: handle -1 step amounts
Wim Taymans [Wed, 18 Jul 2012 09:17:23 +0000 (11:17 +0200)]
basesink: handle -1 step amounts

Define a 0 and -1 step amount. They used to almost do the same thing but now, 0
cancels/stops the current step and -1 keeps on stepping until the end of the
segment.

See https://bugzilla.gnome.org/show_bug.cgi?id=679378

12 years agoquery: fix gst_query_parse_nth_allocation_pool() annotation
Tim-Philipp Müller [Wed, 18 Jul 2012 11:30:59 +0000 (12:30 +0100)]
query: fix gst_query_parse_nth_allocation_pool() annotation

It returns a ref to the pool.

12 years agocheck: Avoid deadlock
Edward Hervey [Tue, 17 Jul 2012 13:52:53 +0000 (15:52 +0200)]
check: Avoid deadlock

Queries will be sent when pipeline goes down to NULL, which would
result in the probe being called ... but can't take the lock.

12 years agogstghostpad: Forward queries in both direction
Edward Hervey [Tue, 17 Jul 2012 13:50:09 +0000 (15:50 +0200)]
gstghostpad: Forward queries in both direction

Use the peer of the internal pad to forward them, instead of the
target which only exists for the ghostpad (and not the internal
proxy pad).

12 years agodocs: More entries
Edward Hervey [Tue, 17 Jul 2012 09:20:43 +0000 (11:20 +0200)]
docs: More entries

12 years agoqueue: answer SCHEDULING query
Tim-Philipp Müller [Wed, 18 Jul 2012 08:15:51 +0000 (09:15 +0100)]
queue: answer SCHEDULING query

Instead of letting the default query handler fail.

12 years agoqueue2: handle CAPS event and drop it if operating in ring buffer mode
Tim-Philipp Müller [Tue, 17 Jul 2012 18:20:23 +0000 (19:20 +0100)]
queue2: handle CAPS event and drop it if operating in ring buffer mode

Fixes "Unexpected event of kind caps can't be added in temp file"
warning when doing download buffering.

12 years agobuffer: make _foreach_meta more powerful
Wim Taymans [Tue, 17 Jul 2012 10:57:59 +0000 (12:57 +0200)]
buffer: make _foreach_meta more powerful

Make _foreach_meta return FALSE when the foreach function returned FALSE.

12 years agobufferlist: improve foreach function
Wim Taymans [Tue, 17 Jul 2012 10:52:59 +0000 (12:52 +0200)]
bufferlist: improve foreach function

Make the foreach function return FALSE when one of the function calls returned
FALSE.

12 years agobuffer: add more debug
Wim Taymans [Tue, 17 Jul 2012 10:50:24 +0000 (12:50 +0200)]
buffer: add more debug

12 years agobasesink: fix debug string
Wim Taymans [Tue, 17 Jul 2012 10:40:54 +0000 (12:40 +0200)]
basesink: fix debug string

12 years agoparse: fix some debug
Wim Taymans [Tue, 17 Jul 2012 07:57:47 +0000 (09:57 +0200)]
parse: fix some debug

12 years agoparse: only escape spaces outside of quotes
Wim Taymans [Tue, 17 Jul 2012 07:48:00 +0000 (09:48 +0200)]
parse: only escape spaces outside of quotes

When we escape spaces to keep arguments together, only escape when the space is
outside a "" string.

See https://bugzilla.gnome.org/show_bug.cgi?id=673319

12 years agoRevert "parse: escape \ with a \ as well, so that we don't lose the \ when unescaping"
Wim Taymans [Tue, 17 Jul 2012 07:44:10 +0000 (09:44 +0200)]
Revert "parse: escape \ with a \ as well, so that we don't lose the \ when unescaping"

This reverts commit dd9fedb41f1ada8e1f8bd5346fccd3d068d543cb.

This is not the right place to escape the \, we should only escape the spaces to
keep the arguments together that were provided as one group (with quotes on the
shell).

12 years agoutils: set return type of gst_parse_bin_* to GstBin for introspection
Evan Nemerson [Tue, 10 Jul 2012 19:27:11 +0000 (12:27 -0700)]
utils: set return type of gst_parse_bin_* to GstBin for introspection

12 years agonettimepacket: add missing array annotation to gst_net_time_packet_new
Evan Nemerson [Sat, 30 Jun 2012 19:33:43 +0000 (12:33 -0700)]
nettimepacket: add missing array annotation to gst_net_time_packet_new

12 years agointrospection: add missing array annotation to gst_formats_contains
Evan Nemerson [Sat, 30 Jun 2012 00:33:49 +0000 (17:33 -0700)]
introspection: add missing array annotation to gst_formats_contains

12 years agobin: aggregate durations like in adder
Stefan Sauer [Mon, 16 Jul 2012 18:54:17 +0000 (20:54 +0200)]
bin: aggregate durations like in adder

Stop querying the duration once an element return unknown and return unknown
as a final result. This avoid eventually cutting off a stream too early.
Add a tests to docuement the behavior.

12 years agodatetime: just return NULL on short input strings instead of a warning
Tim-Philipp Müller [Sun, 15 Jul 2012 23:24:46 +0000 (00:24 +0100)]
datetime: just return NULL on short input strings instead of a warning

We want to be able to use this function on random non-NULL input,
this should not result in a runtime-critical.

12 years agobaseparse: fix seekability querying with formats with headers like FLAC
Tim-Philipp Müller [Sun, 15 Jul 2012 11:59:44 +0000 (12:59 +0100)]
baseparse: fix seekability querying with formats with headers like FLAC

Move code that checks for upstream seekability and all that to
the right place, otherwise it will never be done for formats
that have headers such as FLAC, as handle_and_push frame will
be called the first time only after headers have been processed
(and framecount is > 0). This then makes us report that we
can't seek, which disables the seek bar in totem.

12 years agoplugins: embed GstAueueArray in dataqueue struct as well
Tim-Philipp Müller [Sat, 14 Jul 2012 19:33:30 +0000 (20:33 +0100)]
plugins: embed GstAueueArray in dataqueue struct as well

12 years agoplugins: don't use one-time array in plugin_init
Tim-Philipp Müller [Sat, 14 Jul 2012 19:28:54 +0000 (20:28 +0100)]
plugins: don't use one-time array in plugin_init

12 years agoqueue: embed GstQueueArray structure
Tim-Philipp Müller [Sat, 14 Jul 2012 19:26:04 +0000 (20:26 +0100)]
queue: embed GstQueueArray structure

12 years agoplugins: sprinkle some more G_GNUC_INTERNAL
Tim-Philipp Müller [Sat, 14 Jul 2012 19:00:30 +0000 (20:00 +0100)]
plugins: sprinkle some more G_GNUC_INTERNAL

12 years agoplugins: add init/clear functions to GstQueueArray
Tim-Philipp Müller [Sat, 14 Jul 2012 18:38:39 +0000 (19:38 +0100)]
plugins: add init/clear functions to GstQueueArray

12 years agobase: make GstQueueArray private to coreelements for now
Tim-Philipp Müller [Sat, 14 Jul 2012 18:24:57 +0000 (19:24 +0100)]
base: make GstQueueArray private to coreelements for now

Keep it private until we have a reason to make it public.

12 years agotaglist: check value type matches tag type when adding values to a taglist
Tim-Philipp Müller [Sat, 14 Jul 2012 18:08:24 +0000 (19:08 +0100)]
taglist: check value type matches tag type when adding values to a taglist

12 years agoinfo: make taglists and datetime loggable via GST_PTR_FORMAT
Tim-Philipp Müller [Sat, 14 Jul 2012 17:52:50 +0000 (18:52 +0100)]
info: make taglists and datetime loggable via GST_PTR_FORMAT

12 years agobaseparse: send seek event upstream first
Wim Taymans [Fri, 13 Jul 2012 10:05:15 +0000 (12:05 +0200)]
baseparse: send seek event upstream first

First try to let upstream handle the seek event, then fail if the event is
something we don't understand.

12 years agopad: fix test raciness
Wim Taymans [Fri, 13 Jul 2012 07:43:59 +0000 (09:43 +0200)]
pad: fix test raciness

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=679506

12 years agogst-uninstalled: fix gst-ffmpeg plugin path again
Tim-Philipp Müller [Thu, 12 Jul 2012 12:17:51 +0000 (13:17 +0100)]
gst-uninstalled: fix gst-ffmpeg plugin path again

12 years agogst-uninstalled: add clutter-gst and refine plugin search paths
Tim-Philipp Müller [Thu, 12 Jul 2012 11:09:13 +0000 (12:09 +0100)]
gst-uninstalled: add clutter-gst and refine plugin search paths

12 years agogstpad: Move sticky flag clearing code to gst_pad_activate_mode
Jan Schmidt [Wed, 11 Jul 2012 14:34:22 +0000 (00:34 +1000)]
gstpad: Move sticky flag clearing code to gst_pad_activate_mode

The ghostpad code directly activates/deactivates the child code by
calling gst_pad_activate_mode, rather than gst_pad_set_active, so
make sure to clear the flags in gst_pad_activate_mode(), which should
catch all cases.

12 years agoevent: improve annotation
Wim Taymans [Wed, 11 Jul 2012 10:40:10 +0000 (12:40 +0200)]
event: improve annotation

12 years agobasesink: handle step end correctly
Wim Taymans [Wed, 11 Jul 2012 10:37:05 +0000 (12:37 +0200)]
basesink: handle step end correctly

when we have a new step event with a -1 amount, make sure that we follow the
regular code path so that the stop_end handler is called as usual. This takes
care of flushing the buffer in case of a flushing step and also posts a step end
message.

See https://bugzilla.gnome.org/show_bug.cgi?id=679378

12 years agowin32: Fix exported symbols list for real now
Sebastian Dröge [Wed, 11 Jul 2012 11:14:57 +0000 (13:14 +0200)]
win32: Fix exported symbols list for real now

12 years agosegment: remove removed api from the docs.
Stefan Sauer [Wed, 11 Jul 2012 09:21:18 +0000 (11:21 +0200)]
segment: remove removed api from the docs.

12 years agowin32: Updated exported symbols list
Sebastian Dröge [Wed, 11 Jul 2012 10:46:31 +0000 (12:46 +0200)]
win32: Updated exported symbols list

12 years agotoc: Add functions to retrieve the parent GstToc/GstTocEntry of a GstTocEntry
Sebastian Dröge [Wed, 11 Jul 2012 10:45:51 +0000 (12:45 +0200)]
toc: Add functions to retrieve the parent GstToc/GstTocEntry of a GstTocEntry

12 years agotoc: Fix gst_toc_find_entry()
Anton Belka [Tue, 10 Jul 2012 15:15:20 +0000 (18:15 +0300)]
toc: Fix gst_toc_find_entry()

Recursive search for the required entry, instead of returning the
top-level entry that contains an entry with the search UID.

12 years agobaseparse: Push STREAM_START in pull-mode
Edward Hervey [Wed, 11 Jul 2012 08:26:13 +0000 (10:26 +0200)]
baseparse: Push STREAM_START in pull-mode

12 years agostructure: Demote WARNING to DEBUG
Edward Hervey [Wed, 11 Jul 2012 08:24:51 +0000 (10:24 +0200)]
structure: Demote WARNING to DEBUG

It is not an issue to get fields that don't exist, calling code should
handle that.

12 years agoRemove 0.10-related documentation and "Since" markers
Edward Hervey [Tue, 10 Jul 2012 09:46:41 +0000 (11:46 +0200)]
Remove 0.10-related documentation and "Since" markers

12 years agobasesrc: provide fallback in case a create function doesn't know about provided buffers
Tim-Philipp Müller [Mon, 9 Jul 2012 23:39:37 +0000 (00:39 +0100)]
basesrc: provide fallback in case a create function doesn't know about provided buffers

In 0.11 the caller may provide a buffer to be filled by the source to
pull_range/get_range/create, but it's easy to miss this new case when
porting code from 0.10. Provide fallback that copies the created data
into the provided buffer for now.

This makes oggdemux in pull-mode work with dataurisrc.

12 years agoquery: copy structure in _add_allocation_meta()
Wim Taymans [Tue, 10 Jul 2012 08:31:00 +0000 (10:31 +0200)]
query: copy structure in _add_allocation_meta()

Make gst_query_add_allocation_meta() take a copy of the passed caps instead of
taking ownership. This makes it easier for the caller in most cases because it
doesn't have to make a copy and deal with NULL values.

12 years agogst: add new flags
Wim Taymans [Tue, 10 Jul 2012 08:11:57 +0000 (10:11 +0200)]
gst: add new flags

12 years agominiobject: fix exclusive lock/unlock race
Matej Knopp [Mon, 9 Jul 2012 21:47:53 +0000 (23:47 +0200)]
miniobject: fix exclusive lock/unlock race

12 years agobasesrc, basesink, baseparse, typefind: use GST_SEGMENT_FLAG with segment flags
Tim-Philipp Müller [Mon, 9 Jul 2012 20:51:07 +0000 (21:51 +0100)]
basesrc, basesink, baseparse, typefind: use GST_SEGMENT_FLAG with segment flags

12 years agosegment: also copy the segment flag
Stefan Sauer [Mon, 9 Jul 2012 20:11:31 +0000 (22:11 +0200)]
segment: also copy the segment flag

Fixes segmented seeks (as tested e.g. in the adder tests in base).

12 years agoplugins: sprinkle G_GNUC_INTERNAL for dataqueue functions
Tim-Philipp Müller [Mon, 9 Jul 2012 19:55:00 +0000 (20:55 +0100)]
plugins: sprinkle G_GNUC_INTERNAL for dataqueue functions

And remove padding, since this is not public API any more.

12 years agogst: sprinkle some G_GNUC_INTERNAL for internal functions
Tim-Philipp Müller [Mon, 9 Jul 2012 19:48:56 +0000 (20:48 +0100)]
gst: sprinkle some G_GNUC_INTERNAL for internal functions

12 years agotests: fix toc unit tests
Tim-Philipp Müller [Mon, 9 Jul 2012 19:09:55 +0000 (20:09 +0100)]
tests: fix toc unit tests

Meant to check subsubentry, not subentry.

12 years agotests: minor toc test clean-up
Tim-Philipp Müller [Mon, 9 Jul 2012 17:58:24 +0000 (18:58 +0100)]
tests: minor toc test clean-up

12 years agotests: turn toc check macros into proper functions
Tim-Philipp Müller [Mon, 9 Jul 2012 17:51:37 +0000 (18:51 +0100)]
tests: turn toc check macros into proper functions

So we can see the line number of the check that fails.

12 years agowin32: Update defs file for API changes/addition
Edward Hervey [Mon, 9 Jul 2012 18:31:00 +0000 (20:31 +0200)]
win32: Update defs file for API changes/addition

12 years agocheck: Update tests for new STREAM_START message
Edward Hervey [Mon, 9 Jul 2012 18:29:29 +0000 (20:29 +0200)]
check: Update tests for new STREAM_START message

12 years agocheck: Ensure STREAM_START message is posted
Edward Hervey [Mon, 9 Jul 2012 18:28:54 +0000 (20:28 +0200)]
check: Ensure STREAM_START message is posted

A STREAM_START message is posted if and only if all sinks in the
bin/pipeline received the STREAM_START event

12 years agobasesink: Post a STREAM_START message when we see the event
Edward Hervey [Mon, 9 Jul 2012 18:28:20 +0000 (20:28 +0200)]
basesink: Post a STREAM_START message when we see the event

12 years agogstbin: collect and aggregate STREAM_START messages
Edward Hervey [Mon, 9 Jul 2012 18:27:44 +0000 (20:27 +0200)]
gstbin: collect and aggregate STREAM_START messages

when all sinks have posted a STREAM_START, the bin will forward a
new STREAM_START message to the parent bin or application

12 years agogstmessage: New GST_MESSAGE_STREAM_START
Edward Hervey [Mon, 9 Jul 2012 18:08:15 +0000 (20:08 +0200)]
gstmessage: New GST_MESSAGE_STREAM_START

message counterpart to the GST_EVENT_STREAM_START event

12 years agocheck: Unit test for EOS message
Edward Hervey [Mon, 9 Jul 2012 17:59:33 +0000 (19:59 +0200)]
check: Unit test for EOS message

Make sure we get the aggregated message if and only if all sinks
received an EOS event

12 years agocollectpads: add STREAM_START handling
Stefan Sauer [Mon, 9 Jul 2012 17:56:15 +0000 (19:56 +0200)]
collectpads: add STREAM_START handling

Use a flag to forward the first STREAM_START

12 years agodocs: update stream docs for SEGMENT_START event
Tim-Philipp Müller [Mon, 9 Jul 2012 15:20:49 +0000 (16:20 +0100)]
docs: update stream docs for SEGMENT_START event

12 years agodocs: fix more docs
Wim Taymans [Mon, 9 Jul 2012 14:48:56 +0000 (16:48 +0200)]
docs: fix more docs

12 years agodocs: fix docs a little more
Wim Taymans [Mon, 9 Jul 2012 14:22:04 +0000 (16:22 +0200)]
docs: fix docs a little more

12 years agomemory: Make GstAllocator a GstObject
Wim Taymans [Mon, 9 Jul 2012 14:02:50 +0000 (16:02 +0200)]
memory: Make GstAllocator a GstObject

Make GstAllocator a GstObject instead of a GstMiniObject, like bufferpool.
Make a new gstallocator.c file. Make a GstAllocator subclass for the default
allocator.

12 years agomemory: remove unused macros
Wim Taymans [Mon, 9 Jul 2012 11:20:30 +0000 (13:20 +0200)]
memory: remove unused macros

12 years agoclock: make abstract
Wim Taymans [Mon, 9 Jul 2012 11:20:05 +0000 (13:20 +0200)]
clock: make abstract

Make the GstClock type abstract.
Fix a horrible hack in the clock unit test.

12 years agoqueue: Fix handling of min-threshold and serialized queries
Sebastian Dröge [Mon, 9 Jul 2012 13:37:28 +0000 (15:37 +0200)]
queue: Fix handling of min-threshold and serialized queries

Only consider the queue empty if the minimum thresholds
are not reached and data is at the queue head. Otherwise
we would block forever on serialized queries.

This also makes sending of serialized events, like caps, happen
faster and potentially improves negotiation performance.

Fixes bug #679458.

12 years agotoc: remove padding now that the structs are private
Tim-Philipp Müller [Mon, 9 Jul 2012 12:15:46 +0000 (13:15 +0100)]
toc: remove padding now that the structs are private

12 years agotoc: add gst_toc_dump() function for debugging
Tim-Philipp Müller [Mon, 9 Jul 2012 12:12:27 +0000 (13:12 +0100)]
toc: add gst_toc_dump() function for debugging

API: gst_toc_dump()

12 years agobus, clock: make sure these never have a floating ref
Tim-Philipp Müller [Mon, 2 Jul 2012 23:07:11 +0000 (00:07 +0100)]
bus, clock: make sure these never have a floating ref

Clear the initial floating ref in the init function for
busses and clocks. These objects can be set on multiple
elements, so there's no clear parent-child relationship
here. Ideally we'd just not make them derive from
GInitiallyUnowned at all, but since we want to keep
using GstObject features for debugging, we'll just do
it like this.

This should also fix some problems with bindings, which
seem to get confused when they get floating refs from
non-constructor functions (or functions annotated to
have a 'transfer full' return type). This works now:

from gi.repository import GObject, Gst

GObject.threads_init()
Gst.init(None)

pipeline=Gst.Pipeline()
bus = pipeline.get_bus()
pipeline.set_state(Gst.State.NULL)
del pipeline;

https://bugzilla.gnome.org/show_bug.cgi?id=679286
https://bugzilla.gnome.org/show_bug.cgi?id=657202

12 years agoinspect: suppress glib deprecations warnings for G_VALUE_ARRAY
Stefan Sauer [Sun, 8 Jul 2012 18:15:33 +0000 (20:15 +0200)]
inspect: suppress glib deprecations warnings for G_VALUE_ARRAY

12 years agotests: add more tests for datetime value serialisation/deserialisation
Tim-Philipp Müller [Sat, 7 Jul 2012 22:13:20 +0000 (23:13 +0100)]
tests: add more tests for datetime value serialisation/deserialisation

Esp. of partial datetimes.

12 years agovalue: use datetime serialise/deserialise functions for datetimes
Tim-Philipp Müller [Sat, 7 Jul 2012 21:46:00 +0000 (22:46 +0100)]
value: use datetime serialise/deserialise functions for datetimes

This re-uses existing code and makes sure we properly serialise
and deserialise datetimes where not all fields are set (thus
fixing some warnings when serialising such datetimes).

12 years agodatetime: do our own serialisation so we can serialise microseconds as well
Tim-Philipp Müller [Sat, 7 Jul 2012 21:40:12 +0000 (22:40 +0100)]
datetime: do our own serialisation so we can serialise microseconds as well

We still don't do that in _to_iso8601_string() though, since
this will probably mostly be used in tags, where it doesn't
matter so much and the microsecond argument might not be
well-received by some tag readers.

12 years agodatetime: when deserialising parse microseconds if available
Tim-Philipp Müller [Sat, 7 Jul 2012 18:43:50 +0000 (19:43 +0100)]
datetime: when deserialising parse microseconds if available

12 years agodatetime: fix second parsing failure case when deserialising datetime
Tim-Philipp Müller [Sat, 7 Jul 2012 15:01:41 +0000 (16:01 +0100)]
datetime: fix second parsing failure case when deserialising datetime

When we fail to parse the number of seconds, reset the value to -1
instead of passing some error value as seconds. Also, we can still
try to parse timezone information.

12 years agoexamples: fix debug log print formats in memory examples
Tim-Philipp Müller [Sat, 7 Jul 2012 14:44:57 +0000 (15:44 +0100)]
examples: fix debug log print formats in memory examples

12 years agogstinfo: Add destroy notify arguments to debug stubs
Sebastian Rasmussen [Fri, 6 Jul 2012 23:37:50 +0000 (01:37 +0200)]
gstinfo: Add destroy notify arguments to debug stubs

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=679535

12 years agogststructure: Set lcopy string const exactly as glib's macro
Sebastian Rasmussen [Fri, 6 Jul 2012 18:37:06 +0000 (20:37 +0200)]
gststructure: Set lcopy string const exactly as glib's macro

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=679534

12 years agomemory: expose the GstAllocation structure
Wim Taymans [Fri, 6 Jul 2012 15:19:21 +0000 (17:19 +0200)]
memory: expose the GstAllocation structure

Expose the GstAllocation structure and provide an _init function. This makes it
easier to make 'subclasses' of the allocator that contain more info.
It also allows us to expose the flags on the allocator miniobject.
Make a flag to note that the allocator uses a custom alloc function.

12 years agostructure: Demote WARNING to INFO
Edward Hervey [Fri, 6 Jul 2012 10:45:29 +0000 (12:45 +0200)]
structure: Demote WARNING to INFO

It is common to use gst_structure_get() to know if a field is present
or not.

12 years agogst-inspect: Remove unused define
Edward Hervey [Fri, 6 Jul 2012 09:41:52 +0000 (11:41 +0200)]
gst-inspect: Remove unused define

12 years agocheck: gstcontroller.h doesn't exist anymore
Edward Hervey [Fri, 6 Jul 2012 09:41:33 +0000 (11:41 +0200)]
check: gstcontroller.h doesn't exist anymore

12 years agocheck: Use consistencycheck on basesrc
Edward Hervey [Fri, 6 Jul 2012 09:40:47 +0000 (11:40 +0200)]
check: Use consistencycheck on basesrc