arian/navidrome
1
0
Fork 0
Code Issues Pull Requests 1 Packages Projects Releases Wiki Activity

Add OS Lyrics extension (#2656)

Browse Source 
* draft commit

* time to fight pipeline

* round 2 changes

* remove unnecessary line

* fight taglib. again

* make taglib work again???

* add id3 tags

* taglib 1.12 vs 1.13

* use int instead for windows

* store as json now

* add migration, more tests

* support repeated line, multiline

* fix ms and support .m, .mm, .mmm

* address some concerns, make cpp a bit safer

* separate responses from model

* remove [:]

* Add trace log

* Try to unblock pipeline

* Fix merge errors

* Fix SIGSEGV error (proper handling of empty frames)

* Add fallback artist/title to structured lyrics

* Rename conflicting named vars

* Fix tests

* Do we still need ffmpeg in the pipeline?

* Revert "Do we still need ffmpeg in the pipeline?"

Yes we do.

This reverts commit 87df7f6df79bccee83f48c4b7a8118a7636a5e66.

* Does this passes now, with a newer ffmpeg version?

* Revert "Does this passes now, with a newer ffmpeg version?"

No, it does not :(

This reverts commit 372eb4b0ae05d9ffe98078e9bc4e56a9b2921f32.

* My OCD made me do it :P

---------

Co-authored-by: Deluan Quintão <deluan@navidrome.org>
This commit is contained in:
Kendall Garner
2023-12-28 01:20:29 +00:00
committed by GitHub
parent 130ab76c79
commit 814161d78d
37 changed files with 1215 additions and 71 deletions
Download Patch File Download Diff File Expand all files Collapse all files
scanner/metadata/taglib/taglib_test.go
+51 -13
View File
@@ -39,7 +39,10 @@ var _ = Describe("Extractor", func() {
Expect(m).To(HaveKeyWithValue("album", []string{"Album", "Album"}))
Expect(m).To(HaveKeyWithValue("artist", []string{"Artist", "Artist"}))
Expect(m).To(HaveKeyWithValue("albumartist", []string{"Album Artist"}))
Expect(m).To(HaveKeyWithValue("tcmp", []string{"1"})) // Compilation
Expect(m).To(Or(
HaveKeyWithValue("compilation", []string{"1"}),
HaveKeyWithValue("tcmp", []string{"1"}))) // Compilation
Expect(m).To(HaveKeyWithValue("genre", []string{"Rock"}))
Expect(m).To(HaveKeyWithValue("date", []string{"2014-05-21", "2014"}))
Expect(m).To(HaveKeyWithValue("originaldate", []string{"1996-11-21"}))
@@ -50,7 +53,21 @@ var _ = Describe("Extractor", func() {
Expect(m).To(HaveKeyWithValue("bitrate", []string{"192"}))
Expect(m).To(HaveKeyWithValue("channels", []string{"2"}))
Expect(m).To(HaveKeyWithValue("comment", []string{"Comment1\nComment2"}))
Expect(m).To(HaveKeyWithValue("lyrics", []string{"Lyrics 1\rLyrics 2"}))
Expect(m).ToNot(HaveKey("lyrics"))
Expect(m).To(Or(HaveKeyWithValue("lyrics-eng", []string{
"[00:00.00]This is\n[00:02.50]English SYLT\n",
"[00:00.00]This is\n[00:02.50]English",
}), HaveKeyWithValue("lyrics-eng", []string{
"[00:00.00]This is\n[00:02.50]English",
"[00:00.00]This is\n[00:02.50]English SYLT\n",
})))
Expect(m).To(Or(HaveKeyWithValue("lyrics-xxx", []string{
"[00:00.00]This is\n[00:02.50]unspecified SYLT\n",
"[00:00.00]This is\n[00:02.50]unspecified",
}), HaveKeyWithValue("lyrics-xxx", []string{
"[00:00.00]This is\n[00:02.50]unspecified",
"[00:00.00]This is\n[00:02.50]unspecified SYLT\n",
})))
Expect(m).To(HaveKeyWithValue("bpm", []string{"123"}))
Expect(m).To(HaveKeyWithValue("replaygain_album_gain", []string{"+3.21518 dB"}))
Expect(m).To(HaveKeyWithValue("replaygain_album_peak", []string{"0.9125"}))
@@ -70,10 +87,10 @@ var _ = Describe("Extractor", func() {
// TabLib 1.12 returns 18, previous versions return 39.
// See https://github.com/taglib/taglib/commit/2f238921824741b2cfe6fbfbfc9701d9827ab06b
Expect(m).To(HaveKey("bitrate"))
Expect(m["bitrate"][0]).To(BeElementOf("18", "39", "40", "49"))
Expect(m["bitrate"][0]).To(BeElementOf("18", "39", "40", "43", "49"))
})
DescribeTable("Format-Specific tests",
func(file, duration, channels, albumGain, albumPeak, trackGain, trackPeak string) {
func(file, duration, channels, albumGain, albumPeak, trackGain, trackPeak string, id3Lyrics bool) {
file = "tests/fixtures/" + file
mds, err := e.Parse(file)
Expect(err).NotTo(HaveOccurred())
@@ -113,7 +130,21 @@ var _ = Describe("Extractor", func() {
Expect(m).To(HaveKeyWithValue("channels", []string{channels}))
Expect(m).To(HaveKeyWithValue("comment", []string{"Comment1\nComment2"}))
Expect(m).To(HaveKeyWithValue("lyrics", []string{"Lyrics1\nLyrics 2"}))
if id3Lyrics {
Expect(m).To(HaveKeyWithValue("lyrics-eng", []string{
"[00:00.00]This is\n[00:02.50]English",
}))
Expect(m).To(HaveKeyWithValue("lyrics-xxx", []string{
"[00:00.00]This is\n[00:02.50]unspecified",
}))
} else {
Expect(m).To(HaveKeyWithValue("lyrics", []string{
"[00:00.00]This is\n[00:02.50]unspecified",
"[00:00.00]This is\n[00:02.50]English",
}))
}
Expect(m).To(HaveKeyWithValue("bpm", []string{"123"}))
Expect(m).To(HaveKey("tracknumber"))
@@ -123,25 +154,26 @@ var _ = Describe("Extractor", func() {
},
// ffmpeg -f lavfi -i "sine=frequency=1200:duration=1" test.flac
Entry("correctly parses flac tags", "test.flac", "1.00", "1", "+4.06 dB", "0.12496948", "+4.06 dB", "0.12496948"),
Entry("correctly parses flac tags", "test.flac", "1.00", "1", "+4.06 dB", "0.12496948", "+4.06 dB", "0.12496948", false),
Entry("Correctly parses m4a (aac) gain tags", "01 Invisible (RED) Edit Version.m4a", "1.04", "2", "0.37", "0.48", "0.37", "0.48"),
Entry("Correctly parses m4a (aac) gain tags (uppercase)", "test.m4a", "1.04", "2", "0.37", "0.48", "0.37", "0.48"),
Entry("Correctly parses m4a (aac) gain tags", "01 Invisible (RED) Edit Version.m4a", "1.04", "2", "0.37", "0.48", "0.37", "0.48", false),
Entry("Correctly parses m4a (aac) gain tags (uppercase)", "test.m4a", "1.04", "2", "0.37", "0.48", "0.37", "0.48", false),
Entry("correctly parses ogg (vorbis) tags", "test.ogg", "1.04", "2", "+7.64 dB", "0.11772506", "+7.64 dB", "0.11772506"),
Entry("correctly parses ogg (vorbis) tags", "test.ogg", "1.04", "2", "+7.64 dB", "0.11772506", "+7.64 dB", "0.11772506", false),
// ffmpeg -f lavfi -i "sine=frequency=900:duration=1" test.wma
Entry("correctly parses wma/asf tags", "test.wma", "1.02", "1", "3.27 dB", "0.132914", "3.27 dB", "0.132914"),
// Weird note: for the tag parsing to work, the lyrics are actually stored in the reverse order
Entry("correctly parses wma/asf tags", "test.wma", "1.02", "1", "3.27 dB", "0.132914", "3.27 dB", "0.132914", false),
// ffmpeg -f lavfi -i "sine=frequency=800:duration=1" test.wv
Entry("correctly parses wv (wavpak) tags", "test.wv", "1.00", "1", "3.43 dB", "0.125061", "3.43 dB", "0.125061"),
Entry("correctly parses wv (wavpak) tags", "test.wv", "1.00", "1", "3.43 dB", "0.125061", "3.43 dB", "0.125061", false),
// TODO - these breaks in the pipeline as it uses TabLib 1.11. Once Ubuntu 24.04 is released we can uncomment these tests
// ffmpeg -f lavfi -i "sine=frequency=1000:duration=1" test.wav
//Entry("correctly parses wav tags", "test.wav", "1.00", "1", "3.06 dB", "0.125056", "3.06 dB", "0.125056"),
// Entry("correctly parses wav tags", "test.wav", "1.00", "1", "3.06 dB", "0.125056", "3.06 dB", "0.125056", true),
// ffmpeg -f lavfi -i "sine=frequency=1400:duration=1" test.aiff
//Entry("correctly parses aiff tags", "test.aiff", "1.00", "1", "2.00 dB", "0.124972", "2.00 dB", "0.124972"),
// Entry("correctly parses aiff tags", "test.aiff", "1.00", "1", "2.00 dB", "0.124972", "2.00 dB", "0.124972", true),
)
})
@@ -155,6 +187,12 @@ var _ = Describe("Extractor", func() {
_, err := e.extractMetadata(testFilePath)
Expect(err).To(MatchError(fs.ErrNotExist))
})
It("does not throw a SIGSEGV error when reading a file with an invalid frame", func() {
// File has an empty TDAT frame
md, err := e.extractMetadata("tests/fixtures/invalid-files/test-invalid-frame.mp3")
Expect(err).ToNot(HaveOccurred())
Expect(md).To(HaveKeyWithValue("albumartist", []string{"Elvis Presley"}))
})
})
})
scanner/metadata/taglib/taglib_wrapper.cpp
+80 -5
View File
@@ -3,15 +3,19 @@
#include <typeinfo>
#define TAGLIB_STATIC
#include <aifffile.h>
#include <asffile.h>
#include <fileref.h>
#include <flacfile.h>
#include <id3v2tag.h>
#include <unsynchronizedlyricsframe.h>
#include <synchronizedlyricsframe.h>
#include <mp4file.h>
#include <mpegfile.h>
#include <opusfile.h>
#include <tpropertymap.h>
#include <vorbisfile.h>
#include <wavfile.h>
#include "taglib_wrapper.h"
@@ -58,15 +62,86 @@ int taglib_read(const FILENAME_CHAR_T *filename, unsigned long id) {
}
}
TagLib::ID3v2::Tag *id3Tags = NULL;
// Get some extended/non-standard ID3-only tags (ex: iTunes extended frames)
TagLib::MPEG::File *mp3File(dynamic_cast<TagLib::MPEG::File *>(f.file()));
if (mp3File != NULL) {
if (mp3File->ID3v2Tag()) {
const auto &frameListMap(mp3File->ID3v2Tag()->frameListMap());
id3Tags = mp3File->ID3v2Tag();
}
for (const auto &kv : frameListMap) {
if (!kv.second.isEmpty())
if (id3Tags == NULL) {
TagLib::RIFF::WAV::File *wavFile(dynamic_cast<TagLib::RIFF::WAV::File *>(f.file()));
if (wavFile != NULL && wavFile->hasID3v2Tag()) {
id3Tags = wavFile->ID3v2Tag();
}
}
if (id3Tags == NULL) {
TagLib::RIFF::AIFF::File *aiffFile(dynamic_cast<TagLib::RIFF::AIFF::File *>(f.file()));
if (aiffFile && aiffFile->hasID3v2Tag()) {
id3Tags = aiffFile->tag();
}
}
// Yes, it is possible to have ID3v2 tags in FLAC. However, that can cause problems
// with many players, so they will not be parsed
if (id3Tags != NULL) {
const auto &frames = id3Tags->frameListMap();
for (const auto &kv: frames) {
if (kv.first == "USLT") {
for (const auto &tag: kv.second) {
TagLib::ID3v2::UnsynchronizedLyricsFrame *frame = dynamic_cast<TagLib::ID3v2::UnsynchronizedLyricsFrame *>(tag);
if (frame == NULL) continue;
tags.erase("LYRICS");
const auto bv = frame->language();
char language[4] = {'x', 'x', 'x', '\0'};
if (bv.size() == 3) {
strncpy(language, bv.data(), 3);
}
char *val = (char *)frame->text().toCString(true);
go_map_put_lyrics(id, language, val);
}
} else if (kv.first == "SYLT") {
for (const auto &tag: kv.second) {
TagLib::ID3v2::SynchronizedLyricsFrame *frame = dynamic_cast<TagLib::ID3v2::SynchronizedLyricsFrame *>(tag);
if (frame == NULL) continue;
const auto bv = frame->language();
char language[4] = {'x', 'x', 'x', '\0'};
if (bv.size() == 3) {
strncpy(language, bv.data(), 3);
}
const auto format = frame->timestampFormat();
if (format == TagLib::ID3v2::SynchronizedLyricsFrame::AbsoluteMilliseconds) {
for (const auto &line: frame->synchedText()) {
char *text = (char *)line.text.toCString(true);
go_map_put_lyric_line(id, language, text, line.time);
}
} else if (format == TagLib::ID3v2::SynchronizedLyricsFrame::AbsoluteMpegFrames) {
const int sampleRate = props->sampleRate();
if (sampleRate != 0) {
for (const auto &line: frame->synchedText()) {
const int timeInMs = (line.time * 1000) / sampleRate;
char *text = (char *)line.text.toCString(true);
go_map_put_lyric_line(id, language, text, timeInMs);
}
}
}
}
} else {
if (!kv.second.isEmpty()) {
tags.insert(kv.first, kv.second.front()->toString());
}
}
}
}
@@ -90,7 +165,7 @@ int taglib_read(const FILENAME_CHAR_T *filename, unsigned long id) {
const TagLib::ASF::Tag *asfTags{asfFile->tag()};
const auto itemListMap = asfTags->attributeListMap();
for (const auto item : itemListMap) {
tags.insert(item.first, item.second.front().toString());
tags.insert(item.first, item.second.front().toString());
}
}
scanner/metadata/taglib/taglib_wrapper.go
+33
View File
@@ -103,6 +103,12 @@ func go_map_put_str(id C.ulong, key *C.char, val *C.char) {
do_put_map(id, k, val)
}
//export go_map_put_lyrics
func go_map_put_lyrics(id C.ulong, lang *C.char, val *C.char) {
k := "lyrics-" + strings.ToLower(C.GoString(lang))
do_put_map(id, k, val)
}
func do_put_map(id C.ulong, key string, val *C.char) {
if key == "" {
return
@@ -126,3 +132,30 @@ func go_map_put_int(id C.ulong, key *C.char, val C.int) {
defer C.free(unsafe.Pointer(vp))
go_map_put_str(id, key, vp)
}
//export go_map_put_lyric_line
func go_map_put_lyric_line(id C.ulong, lang *C.char, text *C.char, time C.int) {
language := C.GoString(lang)
line := C.GoString(text)
timeGo := int64(time)
ms := timeGo % 1000
timeGo /= 1000
sec := timeGo % 60
timeGo /= 60
min := timeGo % 60
formatted_line := fmt.Sprintf("[%02d:%02d.%02d]%s\n", min, sec, ms/10, line)
lock.RLock()
defer lock.RUnlock()
key := "lyrics-" + language
m := maps[uint32(id)]
existing, ok := m[key]
if ok {
existing[0] += formatted_line
} else {
m[key] = []string{formatted_line}
}
}
scanner/metadata/taglib/taglib_wrapper.h
+2
View File
@@ -14,6 +14,8 @@ extern "C" {
extern void go_map_put_m4a_str(unsigned long id, char *key, char *val);
extern void go_map_put_str(unsigned long id, char *key, char *val);
extern void go_map_put_int(unsigned long id, char *key, int val);
extern void go_map_put_lyrics(unsigned long id, char *lang, char *val);
extern void go_map_put_lyric_line(unsigned long id, char *lang, char *text, int time);
int taglib_read(const FILENAME_CHAR_T *filename, unsigned long id);
#ifdef __cplusplus