39046bed85
Instance: wrap ClientList in std::unique_ptr<>
2019-05-29 21:29:55 +02:00
71a5c8b819
Main: allocate Instance on the stack
2019-05-29 21:22:25 +02:00
620a39afb4
thread/Slack: use std::chrono::duration
2019-05-29 21:11:32 +02:00
14cee01ba1
Main: simplify Android startup
2019-05-29 15:08:39 +02:00
c782fdb698
Main: use AtScopeExit() to free Android objects
2019-05-29 15:08:38 +02:00
49ba76167e
Main: add overload with ConfigData parameter
2019-05-29 15:08:26 +02:00
93ab957800
Main: convert int return types to void
2019-05-29 14:52:58 +02:00
155c915733
input/buffering: make "mutex" public
2019-05-29 13:44:05 +02:00
971450f0d4
input/InputStream: make IsEOF() and IsAvailable() const
2019-05-29 13:31:54 +02:00
40a48cfba0
PluginUnavailable: add subclass PluginUnconfigured
...
PluginUnconfigured exceptions are logged with level "info" instead of
"error". This suppresses some rather boring messages in the default
log level.
Closes https://github.com/MusicPlayerDaemon/MPD/issues/565
2019-05-23 14:25:14 +02:00
9d1906da8a
{event,output}/Thread: downgrade realtime scheduling errors to "info"
...
These messages can be confusing, but they are not critical.
2019-05-23 14:24:27 +02:00
3d2b180cf8
LogBackend: set default log level to "default"
...
This is the documented value, but for unknown reasons, "info" was
really the default.
This was never noticed because there are only very few "info" level
messages.
2019-05-23 14:14:58 +02:00
f987947730
Log: add Log() and LogFormat() overloads with std::exception_ptr
...
Make LogError()/FormatError() wrappers for those. Now we can log
exceptions with a lower level.
2019-05-23 12:23:28 +02:00
e0d5d88104
Log: make LogLevel the first parameter
...
Prepare for templated functions.
2019-05-23 12:17:59 +02:00
585a745484
Log: use GetFullMessage() to print exceptions
...
Print all nested exceptions on a single line to avoid confusion.
2019-05-22 18:24:45 +02:00
43fe513de8
input/buffering: add API documentation
2019-05-22 12:09:20 +02:00
c1b853ca7c
input/InputStream: define UNKNOWN_SIZE without undefined behavior
2019-05-22 12:01:46 +02:00
4b78038b41
input/InputStream: update API documentation
2019-05-22 12:00:16 +02:00
d651d1abfd
util/SparseBuffer: move #endif to the bottom
2019-05-22 10:43:07 +02:00
5a8b734cfd
tag/Id3Load: include cleanup
2019-05-22 10:23:39 +02:00
31b59a0db6
tag/Generic: allow ScanGenericTags() to throw
...
Propagate the error to the caller instead of logging it.
2019-05-22 10:19:28 +02:00
92f7421715
TagFile: allow ScanFileTags*() to throw
2019-05-22 10:19:27 +02:00
6f1d5105ee
TagStream: allow tag_stream_scan() to throw
2019-05-22 10:19:26 +02:00
9a78371b5c
DetachedSong: allow LoadFile(), Update() to throw
2019-05-22 10:19:25 +02:00
3fc4da382e
db/simple/Song: allow LoadFile(), UpdateFile() to throw
...
Preparing to move logger calls out of lower-level libaries, and
propagating error details to the caller instead.
2019-05-22 10:19:24 +02:00
6ee7d88af0
db/update/Archive: fix inverted nullptr check
...
Regression by commit bbdf2dcf1e
2019-05-22 10:18:18 +02:00
bbdf2dcf1e
db/simple/Song: wrap in std::unique_ptr<>
2019-05-21 22:46:34 +02:00
02bb47dd08
db/simple/Song: move struct Disposer to separate header
...
Allow forward-declaring it.
2019-05-21 22:42:32 +02:00
0c48b8d084
db/simple/Song: add noexcept
2019-05-21 22:41:21 +02:00
8462559b2f
db/simple/Song: include cleanup
2019-05-21 22:40:08 +02:00
319c9699fb
tag/Type: add TAG_GROUPING
...
Map ID3 "TIT1" to this new core tag type.
Closes https://github.com/MusicPlayerDaemon/MPD/issues/563
2019-05-21 22:23:06 +02:00
06a0a4a838
time/Convert: include sys/time.h for struct timeval
...
Closes https://github.com/MusicPlayerDaemon/MPD/issues/562
2019-05-21 10:13:16 +02:00
8942be858b
Merge tag 'v0.21.9'
...
release v0.21.9
2019-05-20 17:23:01 +02:00
66a8fac25e
release v0.21.9
2019-05-20 17:10:58 +02:00
1b902e00b4
doc/protocol.rst: several clarifications
...
Closes https://github.com/MusicPlayerDaemon/MPD/issues/340
2019-05-20 17:06:20 +02:00
45a091c00c
.github: add issue templates
2019-05-20 16:44:02 +02:00
923e66738c
player/Thread: fix "single" mode race condition
...
If the decoder finishes decoding the current song between the two
IsIdle() checks, MPD stops playback instead of starting the decoder
for the next song.
This is usually not visible problem, because the main thread restarts
it via playlist::ResumePlayback(), but that way it, ignores "single"
mode.
As a workaround, this commit adds another "queued" check which
re-enters the player loop and checks again whether to start the
decoder.
Closes https://github.com/MusicPlayerDaemon/MPD/issues/556
2019-05-20 16:22:01 +02:00
ff3e2c0514
player/Thread: remove unnecessary "pipe" check
...
The "queued" flag can only possibly be set if the decoder is still
decoding the current song or if the decoder is stopped. This is also
what the following assert() checks. This check was not necessary.
2019-05-20 16:20:59 +02:00
6922a2f55e
input/buffered: check error in IsAvailable()
2019-05-17 12:43:45 +02:00
219546cb81
input/buffering: check error in IsAvailable()
2019-05-17 12:43:06 +02:00
555a4d738c
input/buffering: pass offset to Read() and eliminate Seek()
...
Another step towards supporting multiple readers.
2019-05-17 12:02:55 +02:00
813567bf5c
input/buffering: use notify_one() to wake up thread
...
There is just one thread.
2019-05-17 11:57:23 +02:00
16a07bc201
input/buffering: remove obsolete thread wakeup
...
The thread will always attempt to read more data since commit
2cf6b77627
2019-05-17 11:56:30 +02:00
1153715608
input/buffering: rename "read_error" to "error"
...
The "seek_error" attribute will be eliminated soon.
2019-05-17 11:29:19 +02:00
b5c7c16fb4
input/buffering: merge multiple exception handlers into RunThread()
2019-05-17 11:23:54 +02:00
302c0515b7
input/buffering: move code to RunThreadLocked()
2019-05-17 11:21:56 +02:00
19e4672a54
input/buffering: use notify_all() instead of notify_one()
...
More preparations to support multiple readers.
2019-05-17 11:17:16 +02:00
c2dd6808e1
input/buffering: make read errors fatal, no recovery
...
If a read error occurs, it is very unlikely that the InputStream will
ever recover. Removing the code removes some code complexity which
just isn't worth it. And it allows supporting multiple readers for
one buffer.
2019-05-17 10:55:58 +02:00
2cf6b77627
input/buffering: eliminate "idle" flag, automatically seek to next hole
2019-05-16 22:42:29 +02:00
a5c09f4ddb
input/buffering: destruct our input while mutex is unlocked
2019-05-16 22:38:37 +02:00
Max Kellermann