While waiting for the input stream to become ready, ignore all
commands except STOP.  This fixes seeking errors with (remote) songs
which the decoder has already finished.

decoder_file_decode() should check for plugin->file_decode, not
plugin->stream_decode().

The stream_decode() and file_decode() methods returned a boolean,
indicating whether they were able to decode the song.  This is
redundant, since we already know that: if decoder_initialized() has
been called (and dc.state==DECODE), the plugin succeeded.  Change both
methods to return void.

Instead of having a seprate try_decode() method, let the
stream_decode() and file_decode() methods decide whether they are able
to decode the song.

When a plugin is unable to decode a song, try the other plugins.

Added lots of assertions to the wrapper functions.

The player did not care about the exact error value, it only checked
whether an error has occured.  This could fit well into
decoder_control.state - introduce a new state "DECODE_STATE_ERROR".

decoder.plugin was a write-only attribute.

Instead of checking the stream_types bit set, we can simply check
whether the methods stream_decode() and file_decode() are implemented.

dc.error wasn't updated when the input stream failed to initialize.

Renamed variables and functions.

If an input stream provides tags (e.g. from an icecast server), send
them in the decoder_data() and decoder_tag() methods.  Removed the
according code from the mp3 and oggvorbis plugins - decoders shouldn't
have to care about stream tags.

This patch also adds the missing decoder_tag() invocation to the mp3
plugin.

The try_decode() method may have read some data from the stream, which
is now lost.  To make this data available to other methods, get it
back by rewinding the input stream after each try_decode() invocation.

The ogg and wavpack plugins did this manually and inconsistently; this
code can now be removed.

Don't pass the "seekable" flag with every decoder_data() invocation.
Since that flag won't change within the file, it is enough to pass it
to decoder_initialized() once per file.

Rename all functions to the new prefix.

No CamelCase in the file name.  The output_buffer struct is going to
be renamed to music_pipe.  There are so many buffer levels in MPD, and
calling this one "output buffer" is wrong, because it's not the last
buffer before the music reaches the output devices.

The decoder_plugin structs must never change.  Don't work with
non-const pointers.

Nearly all mapper functions can fail and will then return NULL.  Add
checks to all callers.

http://xkcd.com/292/

If nobody sent a command, the player isn't waiting for the decoder.
Don't wake it up.

The player should always be woken up when the decoder quits.

Eliminate one goto in decodeStart() by moving some cleanup to
decoder_task().

switch looks much nicer than if/elseif/... and gcc generates nice
warnings when a new command is added to the enum.

Don't return 0/-1 on success/error, but true/false.  Instead of int,
use bool for storing flags.

A decoder_flush() invocation was missing in the FLAC plugin, resulting
in casual assertion failures due to a wrong assumption about the last
chunk's audio format.  It's much easier to remove that decoder_flush()
function and make the decoder thread call ob_flush().

For boolean values and success flags, use bool instead of integer (1/0
for true/false, 0/-1 for success/failure).

Renamed all functions and variables.

There was only one user of pathcpy_trunc(), which can be eliminated.

Provide a struct type which can be forward-declared.  The typedef
InputStream is deprecated now.

Wake up the player as soon as the decoder thread has entered its loop.
This fixes a dead lock when the input is blocking.

The mapper library maps directory and song objects to file system
paths.  With this central library, the code mixture in path.c should
be cleaned up, and we will be able to add neat features like aliasing.

The runtime check suggests that the author has somehow thought
song_get_url(NULL) might be valid.  It should not be.  Replace it with
an assertion.

CamelCase is ugly...  rename all functions.

"bool" should be used in C99 programs for boolean values.

Again, a data type which can be forward-declared.

Why have a "_func" prefix on all method names?  Also don't typedef the
methods, there is no advantage in that.

There was a known deadlocking bug in the notify library: when the
other thread set notify->pending after the according check in
notify_wait(), the latter thread was deadlocked.  Resolve this by
synchronizing all accesses to notify->pending with the notify object's
mutex.  Since notify_signal_sync() was never used, we can remove it.
As a consequence, we don't need notify_enter() and notify_leave()
anymore; eliminate them, too.

Give player.c a better name, meaning that the code is used to control
the player thread.