The interrupt callback will can be called from another thread if the
cache is enabled, and the stream disconnects. Then stream_reconnect()
will call this function from within the cache thread.
mp_input_check_interrupt() is not thread-safe due to read_events() not
being thread-safe. It will call input callbacks added with
mp_input_add_fd() - these callbacks lead to code not protected by locks,
such as reading X11 events.
Solve this by adding a stupid hack, which checks whether the calling
thread is the main playback thread (i.e. calling the input callbacks
will be safe). We can remove this hack later, but it requires at least
moving the VO to its own thread first.
This used global variables for the asynchronous interrupt callback.
Pick the simple and dumb solution and stuff the callback into
mpv_global. Do this because interrupt checking should also work in the
connect phase, and currently stream creation equates connecting.
Ideally, this would be passed to the stream on creation instead, or
connecting would be separated from creation. But since I don't know yet
which is better, and since moving stream/demuxer into their own thread
is something that will happen later, go with the mpv_global solution.
Same change as in e2184fcb, but this time for pull based AOs. This is
slightly controversial, because it will make a fast syscall from e.g.
ao_jack. And according to JackAudio developers, syscalls are evil and
will destroy realtime operation. But I don't think this is an issue at
all.
Still avoid locking a mutex. I'm not sure what jackaudio does in the
worst case - but if they set the jackaudio thread (and only this thread)
to realtime, we might run into deadlock situations due to priority
inversion and such. I'm not quite sure whether this can happen, but I'll
readily follow the cargo cult if it makes hack happy.
This is a regression introduced from moving Lua scripts (including the
OSC) to their own threads. Now OSC and dvdnav can add their bindings at
the same time without coordination, which seems to result in the OSC
winning most time, and thus overriding the dvdnav menu bindings.
Fix this by adding a flag that makes dvdnav menu bindings take priority
over all other bindings.
I don't like this function at all, but it's basically a trick to get the
input's mp_log instance in a case the mp_input_parse_cmd_strv() is
almost certainly not going to output anything. But still make it
somewhat more consistent with mp_input_parse_cmd_strv() - why force the
caller to always use MP_ON_OSD_AUTO?
Until now, there were two functions to add input sources (stuff like
stdin input, slave mode, lirc, joystick). Unify them to a single
function (mp_input_add_fd()), and make sure the associated callbacks
always have a context parameter.
Change the lirc and joystick code such that they take store their state
in a context struct (probably worthless), and use the new mp_msg
replacements (the point of this refactoring).
Additionally, get rid of the ugly USE_FD0_CMD_SELECT etc. ifdeffery in
the terminal handling code.
So you can pass a command as list of strings (each item is an argument),
instead of having to worry about escaping and such.
These functions also take an argument for the default command flags. In
particular, this allows setting saner defaults for commands sent by
program code.
Expose this to Lua as mp.send_commandv command (suggestions for a better
name welcome). The Lua version doesn't allow setting the default command
flags, but it can still use command prefixes. The default flags are
different from input.conf, and disable OSD and property expansion.
Remove these because I'm too lazy to convert them to proper
STREAM_CTRLs. Considering that probably nobody uses radio://, caring
about this is a complete waste of time. I will add these commands back
if someone asks for them, but I don't expect this to happen.
Since m_option.h and options.h are extremely often included, a lot of
files have to be changed.
Moving path.c/h to options/ is a bit questionable, but since this is
mainly about access to config files (which are also handled in
options/), it's probably ok.
Tis drops the silly lib prefixes, and attempts to organize the tree in
a more logical way. Make the top-level directory less cluttered as
well.
Renames the following directories:
libaf -> audio/filter
libao2 -> audio/out
libvo -> video/out
libmpdemux -> demux
Split libmpcodecs:
vf* -> video/filter
vd*, dec_video.* -> video/decode
mp_image*, img_format*, ... -> video/
ad*, dec_audio.* -> audio/decode
libaf/format.* is moved to audio/ - this is similar to how mp_image.*
is located in video/.
Move most top-level .c/.h files to core. (talloc.c/.h is left on top-
level, because it's external.) Park some of the more annoying files
in compat/. Some of these are relicts from the time mplayer used
ffmpeg internals.
sub/ is not split, because it's too much of a mess (subtitle code is
mixed with OSD display and rendering).
Maybe the organization of core is not ideal: it mixes playback core
(like mplayer.c) and utility helpers (like bstr.c/h). Should the need
arise, the playback core will be moved somewhere else, while core
contains all helper and common code.
In input test mode, key bindings won't be executed, but are shown on the
OSD. The OSD includes various information, such as the name of the key,
the command itself, whether it's builtin, and the config file location
it was defined.
The input test mode can be enabled with "--input=test". No effort is
spent trying to react to key bindings that normally exit the player;
they are treated just like any other binding.
If parsing a command fails, its location is printed. The location is
the path to the input.conf, and the line number of the key binding and
the associated input command.
The "no-osd" prefix was introduced earlier to disable OSD selectively
based on the key binding. Extend this, and allow the user to force
display of an OSD bar ("osd-bar"), OSD message ("osd-msg") or both
("osd-msg-bar"). This changes mainly how property setting functions
behave.
The default behavior is still the same.
Allow using the choice type (as it used for command line) for arguments
of input commands. Change the magic integer arguments of some commands
(like seek) to use choices instead. The old numeric values are still
allowed (but only those which made sense before, not arbitrary
integers).
In order to do this, remove the input.c specific types (like
MP_CMD_ARG_INT) completely and specify commands using the m_option
types.
Also, add the special choice "-" to some arguments. It's supposed to
signify the default value, so arguments can be easily skipped. Maybe the
choice option should recognize this and not change the previous value,
but we'll leave this for later.
For now, leave compatibility integer values for all new choice
arguments, e.g. "0" maps to 0. We could let the choice option type do
this automatically, but we don't, because we want user input values and
internal mplayer values decoupled in general. The compatibility options
will be removed one day, too.
Also, remove optional args for strings - would require either annoying
additional code, or copying strings twice. It's not used, so remove it.
If a command is not found, warn about it at loading time (just like
other command parsing errors are printed at loading time).
Add an explicit "ignore" command. input.conf instructs users to use this
command to cancel out existing mapping. This clashed with the
warning added in this commit. Make "ignore" a real command and remove
the specialcasing for it from get_cmd_from_keys(). Now "ignore" is
ignored because it's not handled in command.c.
When input.conf is loaded, verify each command and print a warning if
it's invalid or uses legacy commands. This is done for both the user's
and the embedded config files.
The diff is a bit noisy, because mp_input_parse_cmd() is changed to take
a bstr as argument instead of a char*.
Now it depends on the command whether a property wraps around, or stops
at min/max valid property value.
For practically all properties, it's quite unambiguous what the "switch"
command should have done, and there's technically no need to replace it
with these new commands. More over, most properties that cycle are
boolean anyway. But it seems more orthogonal to make the difference
explicit, rather than hardcoding it. Having different commands also
makes it more explicit to the user what these commands do, both just due
to the naming, and what wrapping policy is used. The code is simpler
too.
Move the code for "switch_ratio" to the M_PROPERTY_SET case of the
"aspect" property. The rules are exactly the same, e.g. setting a ratio
smaller than 0.1 sets the pixel aspect ratio to 1:1. For now, we define
that writing "0" sets the PAR to 1:1, and disallow -1 (possibly reserve
it to reset to default aspect ratio).
osd_show_[property_]text => show_text
osd_show_progression => show_progress
show_text, osd_show_property_text and osd_show_text both map to the
code for the previous osd_show_property_text. The only special thing
about osd_show_text is that you don't need to escape "$". Also,
unfortunately osd_show_property_text requires escaping things twice,
one time for the command parser, and the other time for the property
formatting code, while osd_show_text needed only one level of escaping.
Most input commands had their own policy whether to display an OSD
message for user feedback or not. Some commands had two variants, one
that showed an OSD message and one that didn't (e.g. step_property_osd
and step_property).
Change it such that all commands show a message on the OSD. Add a
"no-osd" modifier that disables OSD for that command. Rename the
"step_property" and "step_property_osd" command to "switch", and rename
"set_property" and "set_property_osd" to "set".
Note that commands which haven't used OSD before still don't use OSD.
That will possibly be fixed later. (E.g. "screenshot" could display an
OSD message instead of just printing a message on the terminal.)
The chapter and edition properties still produce OSD messages even with
"no-osd", because they don't map so well to the property_osd_display[]
mechanism.
There are many input commands which are redundant to properties. They
were parsed like normal commands, but set_property_command() in
command.c handled them automatically using the property mechanism. This
still required having the command specifications around, and the code in
command.c was quite messy.
Replace this with a text based replacement mechanism. Some corner cases
are not handled: commands of form "seek_chapter 3 1" are supposed to set
the "chapter" property to 3. This use is probably rare, and doesn't show
up in the default input.conf.
The reason compatibility is kept is because breaking input.conf is quite
annoying, so a minimal effort is made to avoid this. Currently we print
an annoying warning every time a legacy command is used, though.
Also add a compatibility entry for "pt_step", which was removed some
time ago. Variations in whitespace are not handled, but it's good enough
to deal with old input.conf entries.
These have been replaced by properties. Also remove some other slave-
mode specific get commands that can be replaced by property uses.
The get_metadata() function didn't actually contain anything useful,
and just replicated code from other parts of mplayer.
Introduce a general track struct for every audio/video/subtitle track
known to the frontend. External files (subtitles) are now represented
as tracks too. This mainly serves to clean up the subtitle selection
code: now every subtitle is simply a track, instead of using a messy
numbering that goes by subtitle type (as it was stored in the
global_sub_pos field). The mplayer fontend will list external subtitle
files as additional tracks.
The timeline code now tries to match the exact demuxer IDs of all
tracks. This may cause problems when Matroska files with different
track numberings are used with EDL timelines. Change demux_lavf not
to set demuxer IDs, since most time they are not set.
This messes deeply with the subtitle bookkeeping data structures, and
would have to be reimplemented anyway.
It's not sure what this was even useful for. Possibly for slave mode.
Add a flags parameter to mp_input_set_section(). Add a flag that defines
whether bindings in the default section are used or not. This is useful
for special functionality, where the normal key bindings may have
unwanted effects.
For example, it shouldn't be possible to seek during encoding. However,
you want to be able to cancel the encoding process gracefully. For that
purpose, the "encode" section of input.conf could be made exclusive:
mp_input_set_section(mpctx->input, "encode", MP_INPUT_NO_DEFAULT_SECTION);
And input.conf could contain this definition:
RIGHT seek 10
q {encode} quit
Then only the key "q" would be bound during encoding.
When the internal mplayer MPEG demuxer was removed (commit 1fde09db),
the default demuxer when using dvdnav was set to libavformat. Now it
turns out that this doesn't work with libavformat. It will terminate
playback right after the audio runs out (instead of looping it like the
video, or whatever it's supposed to do). I'm not sure what exactly the
problem is, but since 1. even mplayer-svn can't handle DVD menus
directly (missing highlights), 2. DVD menus are essentially worthless,
and 3. I don't directly watch DVDs, don't bother with it and remove it.
For basic playback, there's still libdvdread support.
Also, use pkg-config for libdvdread, and drop support for in-tree
libdvdread. Remove support for in-tree libdvdcss as well.
Teletext requires special OSD support. Because I can't even test
teletext, I can't restore support for it. Since teletext can be
considered ancient and obscure, and since it doesn't make sense to keep
the remaining teletext code without being able to use it, I'm removing
it.
While this was an interesting idea, it wasn't actually useful.
Basically it dumped the raw data (as requested by the demuxer) into a
file. The result is only useful if the file format was raw or maybe
some MPEG packet stream, but not with most modern file formats.
This was done with the help of callcatcher [1]. Only functions which
are statically known to be unused are removed.
Some unused functions are not removed yet, because they might be needed
in the near future (such as open_output_stream for the encode branch).
There is one user visible change: the --subcc option did nothing, and is
removed with this commit.
[1] http://www.skynet.ie/~caolan/Packages/callcatcher.html
Summary:
- There is no playtree anymore. It's reduced to a simple list.
- Options are now always global. You can still have per-file options,
but these are optional and require special syntax.
- The slave command pt_step has been removed, and playlist_next
and playlist_prev added. (See etc/input.conf changes.)
This is a user visible incompatible change, and will break slave-mode
applications.
- The pt_clear slave command is renamed to playlist_clear.
- Playtree entries could have multiple files. This is not the case
anymore, and playlist entries have always exactly one entry. Whenever
something adds more than one file (like ASX playlists or dvd:// or
dvdnav:// on the command line), all files are added as separate
playlist entries.
Note that some of the changes are quite deep and violent. Expect
regressions.
The playlist parsing code in particular is of low quality. I didn't try
to improve it, and merely spent to least effort necessary to keep it
somehow working. (Especially ASX playlist handling.)
The playtree code was complicated and bloated. It was also barely used.
Most users don't even know that mplayer manages the playlist as tree,
or how to use it. The most obscure features was probably specifying a
tree on command line (with '{' and '}' to create/close tree nodes). It
filled the player code with complexity and confused users with weird
slave commands like pt_up.
Replace the playtree with a simple flat playlist. Playlist parsers that
actually return trees are changed to append all files to the playlist
pre-order.
It used to be the responsibility of the playtree code to change per-file
config options. Now this is done by the player core, and the playlist
code is free of such details.
Options are not per-file by default anymore. This was a very obscure and
complicated feature that confused even experienced users. Consider the
following command line:
mplayer file1.mkv file2.mkv --no-audio file3.mkv
This will disable the audio for file2.mkv only, because options are
per-file by default. To make the option affect all files, you're
supposed to put it before the first file.
This is bad, because normally you don't need per-file options. They are
very rarely needed, and the only reasonable use cases I can imagine are
use of the encode backend (mplayer encode branch), or for debugging. The
normal use case is made harder, and the feature is perceived as bug.
Even worse, correct usage is hard to explain for users.
Make all options global by default. The position of an option isn't
significant anymore (except for options that compensate each other,
consider --shuffle --no-shuffle).
One other important change is that no options are reset anymore if a
new file is started. If you change settings with slave mode commands,
they will not be changed by playing a new file. (Exceptions include
settings that are too file specific, like audio/subtitle stream
selection.)
There is still some need for per-file options. Debugging and encoding
are use cases that profit from per-file options. Per-file profiles (as
well as per-protocol and per-VO/AO options) need the implementation
related mechanisms to backup and restore options when the playback file
changes.
Simplify the save-slot stuff, which is possible because there is no
hierarchical play tree anymore. Now there's a simple backup field.
Add a way to specify per-file options on command line. Example:
mplayer f1.mkv -o0 --{ -o1 f2.mkv -o2 f3.mkv --} f4.mkv -o3
will have the following options per file set:
f1.mkv, f4.mkv: -o0 -o3
f2.mkv, f3.mkv: -o0 -o3 -o1 -o2
The options --{ and --} start and end per-file options. All files inside
the { } will be affected by the options equally (similar to how global
options and multiple files are handled). When playback of a file starts,
the per-file options are set according to the command line. When
playback ends, the per-file options are restored to the values when
playback started.
The command lists the audio and subtitle tracks in the current file on the
OSD. It also marks the currently active streams.
Video streams are not shown, as files with more than one video stream are
exceedingly rare.
The command lists the chapters in the current file on the OSD. It also
marks the current chapter.
This is actually a cheap replacement for the chapter select libmenu
functionality.
For ao_pulse, the current latency is not a good indicator of how soon
the AO requires new data to avoid underflow. Add an internal pipe that
can be used to wake up the input loop from select(), and make the
pulseaudio main loop (which runs in a separate thread) use this
mechanism when pulse requests more data. The wakeup signal currently
contains no information about the reason for the wakup, but audio
buffers are always filled when the event loop wakes up.
Also, request a latency of 1 second from the Pulseaudio server. The
default is normally significantly higher. We don't need low latency,
while higher latency helps prevent underflows reduces need for
wakeups.
This deletes all playlist elements, except the currently active
playlist entry.
NOTE: this doesn't remove parent nodes in the case when we really have
a play tree (as opposed to a play list). Apparently this doesn't cause
any harm.
Using type = -1 for terminating the command argument list was an
unnecessary complication. Change it to 0 to make use of default
initialization. Remove the explicit termination from the command
definitions. Now the argument list is automatically terminated by C
default initialization rules.
Also remove other redundant {0} initializers.
When an argument's default value is actually initialized to something other
than 0, make the field being initialized explicit (e.g. {1} becomes
{.i = 1}).
Try to make use of whitespace more consistent.
Note: now that libmenu has been removed, the related code in input.c has
been removed. If libmenu is ever added back, a queue_remove() function
that can remove any item in the input queue will be required.
libmenu triggered some special case that caused crashes by double frees.
In particular the queue_pop function is used, expecting to remove an element
that is not necessarily always the one that queue_pop removes. Fix by
adding a queue_remove functions. Rewrite the other queue functions, since
these are probably buggy as well. queue_pop didn't even attempt to
maintain the doubly linked list.
