| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
| |
We are going to split the pcm_utils.c library, and pcm_range() will be
useful for several sub libraries.
|
|
|
|
| |
Get rid of the non-portable Linux list library, part III (final).
|
|
|
|
| |
Get rid of the non-portable Linux list library, part II.
|
|
|
|
| |
Get rid of the non-portable Linux list library, part I.
|
|
|
|
| |
Fix a typo in the icy-metadata patch.
|
|
|
|
|
| |
Declare and initialize MPD_LIBS and MPD_CFLAGS before the "OS specific
defaults" section.
|
|
|
|
| |
Both are unused.
|
|
|
|
|
| |
Yesterday's patch set changed tagItemFoundAndMatches() and moved the
strcmp() return value into a bool, but forgot to add the ==0 check.
|
|
|
|
| |
"#ifdef G_BYTE_ORDER == G_BIG_ENDIAN" cannot work, of course.
|
|
|
|
|
|
| |
Zeroconf, curl, libsamplerate. No libid3tag for now, because
libid3tag should be automatically disabled when no decoders using it
are enabled.
|
|
|
|
|
| |
Don't separate basic options from their according tests. Due to lots
of interdependencies, we won't do that for the plugins yet.
|
|
|
|
|
| |
Print input plugins first, then archive, metadatab, decoder,
converter, encoder, output plugins.
|
|
|
|
|
| |
Initialize the CFLAGS (warnings, errors) after all the libraries,
because some library checks may be broken with -Werror or -pedantic.
|
|
|
|
|
| |
AC_CHECK_HEADERS defines HAVE_LOCALE_H, so we don't have to manually
define HAVE_LOCALE.
|
|
|
|
|
| |
First do the checks for mandatory stuff (libc features, glib,
pkg-config), then all the options.
|
|
|
|
| |
Don't check for inttypes.h and langinfo.h.
|
| |
|
|
|
|
| |
Fix a typo.
|
|
|
|
| |
Use g_strndup() instead of g_memdup().
|
|
|
|
| |
Make db_load(), db_save() and db_check() return bool instead of int.
|
|
|
|
|
| |
For updating the database, directory_update_init() should be called
explicitly.
|
| |
|
|
|
|
| |
It's unused.
|
|
|
|
| |
The notify library is easier to use, and has no disadvantages.
|
| |
|
|
|
|
|
| |
Start the daemon after --create-db. This makes --create-db a flag
which discards the old database and starts with a fresh one.
|
| |
|
|
|
|
|
|
| |
Handle the DELETE and UPDATE events in separate callbacks:
song_delete_event() safely deletes a song, and update_finished_event()
is called when database update is complete.
|
|
|
|
|
| |
GLib mandates that you initialize all GError objects with NULL prior
to passing it.
|
|
|
|
|
| |
Pass NULL instead of &error to g_convert(). We're not interested in
the error object.
|
|
|
|
| |
Use GLib's g_build_filename() instead of pfx_dir().
|
| |
|
| |
|
|
|
|
|
| |
directory_is_root() is cheaper than
isRootDirectory(directory_get_path()).
|
| |
|
|
|
|
|
|
| |
Don't call command_error() if loading a song from the playlist fails.
This may result in assertion failures, since command_error() may be
called more than once.
|
| |
|
| |
|
|
|
|
| |
The struct delete_data has only one member left and can be eliminated.
|
|
|
|
|
|
|
|
| |
Determine the suffix manually, and use decoder_plugin_from_suffix()
and archive_plugin_from_suffix() instead.
This way, song_file_update_inarchive() can be optimized: it does not
have to translate its path.
|
| |
|
|
|
|
| |
Newline characters are already checked in skip_path() (update.c).
|
|
|
|
|
| |
Don't include headers which are not used. Fix some includes in
decoder_thread.c.
|
|
|
|
|
| |
Reverse the condition: delete directories which don't exist anymore.
This typo caused a slowdown during partial database update.
|
|
|
|
|
|
| |
For internal checks (i.e. not in command.c), we need to check whether
an URI is in the databse, in the local file system or a remote URI
with a scheme.
|
|
|
|
|
| |
Check if the URI scheme is supported by MPD, and print an error
message if not. Optimize the checks in "add" and "playlistadd".
|
|
|
|
| |
Don't work with argv[1], give it the better name "uri".
|
|
|
|
|
|
| |
libid3tag comes without a pkg-config file, and it is usually added by
distribution packages. For those without .pc file, attempt to
auto-detect the library with AC_CHECK_LIB.
|
|
|
|
|
|
|
|
| |
When the decoder of the new song is not fast enough, the player thread
has to wait for it for a moment. However the variable "nextChunk" was
reset to -1 during that, making the next loop iteration assume that
cross-fading has not begun yet. This patch overwrites it with "0"
while waiting.
|
| |
|