1
|
|
|
<?php declare(strict_types=1); |
2
|
|
|
|
3
|
|
|
/** |
4
|
|
|
* ownCloud - Music app |
5
|
|
|
* |
6
|
|
|
* This file is licensed under the Affero General Public License version 3 or |
7
|
|
|
* later. See the COPYING file. |
8
|
|
|
* |
9
|
|
|
* @author Pauli Järvinen <[email protected]> |
10
|
|
|
* @copyright Pauli Järvinen 2018 - 2025 |
11
|
|
|
*/ |
12
|
|
|
|
13
|
|
|
namespace OCA\Music\Service; |
14
|
|
|
|
15
|
|
|
use OCP\Files\File; |
16
|
|
|
use OCP\Files\Folder; |
17
|
|
|
|
18
|
|
|
use OCA\Music\AppFramework\Core\Logger; |
19
|
|
|
use OCA\Music\Utility\StringUtil; |
20
|
|
|
|
21
|
|
|
class DetailsService { |
22
|
|
|
private Extractor $extractor; |
23
|
|
|
private Logger $logger; |
24
|
|
|
|
25
|
|
|
public function __construct( |
26
|
|
|
Extractor $extractor, |
27
|
|
|
Logger $logger) { |
28
|
|
|
$this->extractor = $extractor; |
29
|
|
|
$this->logger = $logger; |
30
|
|
|
} |
31
|
|
|
|
32
|
|
|
public function getDetails(int $fileId, Folder $userFolder) : ?array { |
33
|
|
|
$file = $userFolder->getById($fileId)[0] ?? null; |
34
|
|
|
if ($file instanceof File) { |
35
|
|
|
$data = $this->extractor->extract($file); |
36
|
|
|
$audio = $data['audio'] ?? []; |
37
|
|
|
$comments = $data['comments'] ?? []; |
38
|
|
|
|
39
|
|
|
// remove intermediate arrays |
40
|
|
|
$comments = self::flattenComments($comments); |
41
|
|
|
|
42
|
|
|
// cleanup strings from invalid characters |
43
|
|
|
\array_walk($audio, [$this, 'sanitizeString']); |
44
|
|
|
\array_walk($comments, [$this, 'sanitizeString']); |
45
|
|
|
|
46
|
|
|
$result = [ |
47
|
|
|
'fileinfo' => $audio, |
48
|
|
|
'tags' => $comments |
49
|
|
|
]; |
50
|
|
|
|
51
|
|
|
// binary data has to be encoded |
52
|
|
|
if (\array_key_exists('picture', $result['tags'])) { |
53
|
|
|
$result['tags']['picture'] = self::encodePictureTag($result['tags']['picture']); |
54
|
|
|
} |
55
|
|
|
|
56
|
|
|
// 'streams' contains duplicate data |
57
|
|
|
unset($result['fileinfo']['streams']); |
58
|
|
|
|
59
|
|
|
// one track number is enough |
60
|
|
|
if (\array_key_exists('track', $result['tags']) |
61
|
|
|
&& \array_key_exists('track_number', $result['tags'])) { |
62
|
|
|
unset($result['tags']['track']); |
63
|
|
|
} |
64
|
|
|
|
65
|
|
|
// special handling for lyrics tags |
66
|
|
|
$lyricsNode = self::transformLyrics($result['tags'], self::getLyricsFileContent($file)); |
67
|
|
|
if ($lyricsNode !== null) { |
68
|
|
|
$result['lyrics'] = $lyricsNode; |
69
|
|
|
unset( |
70
|
|
|
$result['tags']['LYRICS'], |
71
|
|
|
$result['tags']['lyrics'], |
72
|
|
|
$result['tags']['unsynchronised_lyric'], |
73
|
|
|
$result['tags']['unsynced lyrics'], |
74
|
|
|
$result['tags']['unsynced_lyrics'], |
75
|
|
|
$result['tags']['unsyncedlyrics'] |
76
|
|
|
); |
77
|
|
|
} |
78
|
|
|
|
79
|
|
|
// add track length |
80
|
|
|
$result['length'] = $data['playtime_seconds'] ?? null; |
81
|
|
|
|
82
|
|
|
// add file path |
83
|
|
|
$result['path'] = $userFolder->getRelativePath($file->getPath()); |
84
|
|
|
|
85
|
|
|
return $result; |
86
|
|
|
} |
87
|
|
|
return null; |
88
|
|
|
} |
89
|
|
|
|
90
|
|
|
/** |
91
|
|
|
* Check if a file has embedded lyrics without parsing them |
92
|
|
|
*/ |
93
|
|
|
public function hasLyrics(int $fileId, Folder $userFolder) : bool { |
94
|
|
|
$fileNode = $userFolder->getById($fileId)[0] ?? null; |
95
|
|
|
if ($fileNode instanceof File) { |
96
|
|
|
$data = $this->extractor->extract($fileNode); |
97
|
|
|
$lyrics = ExtractorGetID3::getFirstOfTags($data, ['unsynchronised_lyric', 'unsynced lyrics', 'unsynced_lyrics', 'unsyncedlyrics']) |
98
|
|
|
?? self::getLyricsFileContent($fileNode); |
99
|
|
|
return ($lyrics !== null); |
100
|
|
|
} |
101
|
|
|
return false; |
102
|
|
|
} |
103
|
|
|
|
104
|
|
|
/** |
105
|
|
|
* Get lyrics from the file metadata in plain-text format. If there's no unsynchronised lyrics available |
106
|
|
|
* but there is synchronised lyrics, then the plain-text format is converted from the synchronised lyrics. |
107
|
|
|
*/ |
108
|
|
|
public function getLyricsAsPlainText(int $fileId, Folder $userFolder) : ?string { |
109
|
|
|
$lyrics = null; |
110
|
|
|
$fileNode = $userFolder->getById($fileId)[0] ?? null; |
111
|
|
|
if ($fileNode instanceof File) { |
112
|
|
|
$data = $this->extractor->extract($fileNode); |
113
|
|
|
$lyrics = ExtractorGetID3::getFirstOfTags($data, ['unsynchronised_lyric', 'unsynced lyrics', 'unsynced_lyrics', 'unsyncedlyrics']); |
114
|
|
|
self::sanitizeString($lyrics); |
115
|
|
|
|
116
|
|
|
if ($lyrics === null) { |
117
|
|
|
// no unsynchronized lyrics, try to get and convert the potentially synchronized lyrics |
118
|
|
|
$lyrics = ExtractorGetID3::getFirstOfTags($data, ['LYRICS', 'lyrics']) ?? self::getLyricsFileContent($fileNode); |
119
|
|
|
self::sanitizeString($lyrics); |
120
|
|
|
$parsed = LyricsParser::parseSyncedLyrics($lyrics); |
121
|
|
|
if ($parsed) { |
122
|
|
|
// the lyrics were indeed time-synced, convert the parsed array to a plain string |
123
|
|
|
$lyrics = LyricsParser::syncedToUnsynced($parsed); |
124
|
|
|
} |
125
|
|
|
} |
126
|
|
|
} |
127
|
|
|
return $lyrics; |
128
|
|
|
} |
129
|
|
|
|
130
|
|
|
/** |
131
|
|
|
* Get all lyrics from the file metadata, both synced and unsynced. For both lyrics types, a single instance |
132
|
|
|
* of lyrics contains an array of lyrics lines. In case of synced lyrics, the key of this array is the offset |
133
|
|
|
* in milliseconds. |
134
|
|
|
* @return array of items like ['synced' => boolean, 'lines' => array] |
135
|
|
|
*/ |
136
|
|
|
public function getLyricsAsStructured(int $fileId, Folder $userFolder) : array { |
137
|
|
|
$result = []; |
138
|
|
|
$fileNode = $userFolder->getById($fileId)[0] ?? null; |
139
|
|
|
if ($fileNode instanceof File) { |
140
|
|
|
$data = $this->extractor->extract($fileNode); |
141
|
|
|
$lyricsTags = ExtractorGetID3::getTags($data, ['LYRICS', 'lyrics', 'unsynchronised_lyric', 'unsynced lyrics', 'unsynced_lyrics', 'unsyncedlyrics']); |
142
|
|
|
|
143
|
|
|
$lrcFileContent = self::getLyricsFileContent($fileNode); |
144
|
|
|
if ($lrcFileContent !== null) { |
145
|
|
|
// prepend the LRC file content to make it the preferred option for the clients |
146
|
|
|
$lyricsTags = \array_merge(['lyrics_file' => $lrcFileContent], $lyricsTags); |
147
|
|
|
} |
148
|
|
|
|
149
|
|
|
foreach ($lyricsTags as $tagKey => $tagValue) { |
150
|
|
|
self::sanitizeString($tagValue); |
151
|
|
|
|
152
|
|
|
// Never try to parse synced lyrics from the "unsync*" tags. The "lyrics" tag, on the other hand, |
153
|
|
|
// may contain either synced or unsynced lyrics and a parse attempt is needed to find out. |
154
|
|
|
$mayBeSynced = !StringUtil::startsWith($tagKey, 'unsync'); |
155
|
|
|
$syncedLyrics = $mayBeSynced ? LyricsParser::parseSyncedLyrics($tagValue) : null; |
156
|
|
|
|
157
|
|
|
if ($syncedLyrics) { |
158
|
|
|
$result[] = [ |
159
|
|
|
'synced' => true, |
160
|
|
|
'lines' => $syncedLyrics |
161
|
|
|
]; |
162
|
|
|
} else { |
163
|
|
|
$result[] = [ |
164
|
|
|
'synced' => false, |
165
|
|
|
'lines' => \explode("\n", $tagValue) |
166
|
|
|
]; |
167
|
|
|
} |
168
|
|
|
} |
169
|
|
|
} |
170
|
|
|
return $result; |
171
|
|
|
} |
172
|
|
|
|
173
|
|
|
/** |
174
|
|
|
* Read lyrics-related tags, and build a result array containing potentially |
175
|
|
|
* both time-synced and unsynced lyrics. If no lyrics tags are found, the result will |
176
|
|
|
* be null. In case the result is non-null, there is always at least the key 'unsynced' |
177
|
|
|
* in the result which will hold a string representing the lyrics with no timestamps. |
178
|
|
|
* If found and successfully parsed, there will be also another key 'synced', which will |
179
|
|
|
* hold the time-synced lyrics. These are presented as an array of arrays of form |
180
|
|
|
* ['time' => int (ms), 'text' => string]. |
181
|
|
|
*/ |
182
|
|
|
private static function transformLyrics(array $tags, ?string $lrcFileContent) : ?array { |
183
|
|
|
$lyrics = $tags['LYRICS'] ?? $tags['lyrics'] ?? null; // may be synced or unsynced |
184
|
|
|
$syncedLyrics = LyricsParser::parseSyncedLyrics($lrcFileContent) ?? LyricsParser::parseSyncedLyrics($lyrics); |
185
|
|
|
$unsyncedLyrics = $tags['unsynchronised_lyric'] |
186
|
|
|
?? $tags['unsynced lyrics'] |
187
|
|
|
?? $tags['unsynced_lyrics'] |
188
|
|
|
?? $tags['unsyncedlyrics'] |
189
|
|
|
?? LyricsParser::syncedToUnsynced($syncedLyrics) |
190
|
|
|
?? $lrcFileContent |
191
|
|
|
?? $lyrics; |
192
|
|
|
|
193
|
|
|
if ($unsyncedLyrics !== null) { |
194
|
|
|
$result = ['unsynced' => $unsyncedLyrics]; |
195
|
|
|
|
196
|
|
|
if ($syncedLyrics !== null) { |
197
|
|
|
$result['synced'] = \array_map(fn($timestamp, $text) => [ |
198
|
|
|
'time' => \max(0, $timestamp), 'text' => $text |
199
|
|
|
], \array_keys($syncedLyrics), $syncedLyrics); |
200
|
|
|
} |
201
|
|
|
} else { |
202
|
|
|
$result = null; |
203
|
|
|
} |
204
|
|
|
|
205
|
|
|
return $result; |
206
|
|
|
} |
207
|
|
|
|
208
|
|
|
private static function getLyricsFileContent(File $audioFile) : ?string { |
209
|
|
|
$audioName = $audioFile->getName(); |
210
|
|
|
$bareName = \pathinfo($audioName, PATHINFO_FILENAME); |
211
|
|
|
$parentDir = $audioFile->getParent(); |
212
|
|
|
|
213
|
|
|
// if the audio file is named "someSong.mp3", we allow the lyrics file to be named "someSong.lrc" or "someSong.mp3.lrc" |
214
|
|
|
$lrcName1 = "$bareName.lrc"; |
215
|
|
|
$lrcName2 = "$audioName.lrc"; |
216
|
|
|
|
217
|
|
|
if ($parentDir->nodeExists($lrcName1)) { |
218
|
|
|
$lrcFile = $parentDir->get($lrcName1); |
219
|
|
|
} else if ($parentDir->nodeExists($lrcName2)) { |
220
|
|
|
$lrcFile = $parentDir->get($lrcName2); |
221
|
|
|
} |
222
|
|
|
|
223
|
|
|
return (isset($lrcFile) && $lrcFile instanceof File) ? $lrcFile->getContent() : null; |
224
|
|
|
} |
225
|
|
|
|
226
|
|
|
/** |
227
|
|
|
* Base64 encode the picture binary data and wrap it so that it can be directly used as |
228
|
|
|
* src of an HTML img element. |
229
|
|
|
*/ |
230
|
|
|
private static function encodePictureTag(array $pic) : ?string { |
231
|
|
|
if ($pic['data']) { |
232
|
|
|
return 'data:' . $pic['image_mime'] . ';base64,' . \base64_encode($pic['data']); |
233
|
|
|
} else { |
234
|
|
|
return null; |
235
|
|
|
} |
236
|
|
|
} |
237
|
|
|
|
238
|
|
|
/** |
239
|
|
|
* Remove potentially invalid characters from the string and normalize the line breaks to LF. |
240
|
|
|
* @param mixed $item |
241
|
|
|
*/ |
242
|
|
|
private static function sanitizeString(&$item) : void { |
243
|
|
|
if (\is_string($item)) { |
244
|
|
|
\mb_substitute_character(0xFFFD); // Use the Unicode REPLACEMENT CHARACTER (U+FFFD) |
245
|
|
|
$item = \mb_convert_encoding($item, 'UTF-8', 'UTF-8'); |
246
|
|
|
// The tags could contain line breaks in formats LF, CRLF, or CR, but we want the output |
247
|
|
|
// to always use the LF style. Note that the order of the next two lines is important! |
248
|
|
|
$item = \str_replace("\r\n", "\n", $item); |
249
|
|
|
$item = \str_replace("\r", "\n", $item); |
250
|
|
|
} |
251
|
|
|
} |
252
|
|
|
|
253
|
|
|
/** |
254
|
|
|
* In the 'comments' field from the extractor, the value for each key is a 1-element |
255
|
|
|
* array containing the actual tag value. Remove these intermediate arrays. |
256
|
|
|
*/ |
257
|
|
|
private static function flattenComments(array $array) : array { |
258
|
|
|
// key 'text' is an exception, its value is an associative array |
259
|
|
|
$textArray = null; |
260
|
|
|
|
261
|
|
|
foreach ($array as $key => $value) { |
262
|
|
|
if ($key === 'text') { |
263
|
|
|
$textArray = $value; |
264
|
|
|
} else { |
265
|
|
|
$array[$key] = $value[0]; |
266
|
|
|
} |
267
|
|
|
} |
268
|
|
|
|
269
|
|
|
if (!empty($textArray)) { |
270
|
|
|
$array = \array_merge($array, $textArray); |
271
|
|
|
unset($array['text']); |
272
|
|
|
} |
273
|
|
|
|
274
|
|
|
return $array; |
275
|
|
|
} |
276
|
|
|
} |
277
|
|
|
|