Write lyrics to track metadata (#194)

* TODO: Don't throw traceback if no lyrics found

* Add more metadata fields

* Refactor debug logging

* Fix traceback when lyrics not found

* It already vomits metadata :3

* Bump lyricwikia >= 0.1.8

README.md

@@ -7,12 +7,13 @@
 
 - Can also download a song by entering its artist and song name (in case if you don't have the Spotify's HTTP link for some song).
 
-- Automatically fixes song's meta-tags which include:
+- Automatically applies metadata to the downloaded song which include:
 
   - Title
   - Artist
   - Album
   - Album art
+  - Lyrics (if found on http://lyrics.wikia.com)
   - Album artist
   - Genre
   - Track number

core/metadata.py

@@ -1,5 +1,5 @@
 from mutagen.easyid3 import EasyID3
-from mutagen.id3 import ID3, APIC
+from mutagen.id3 import ID3, TORY, TYER, TPUB, APIC, USLT, COMM
 from mutagen.mp4 import MP4, MP4Cover
 from core.logger import log
 
@@ -43,6 +43,9 @@ def embed(music_file, meta_tags):
 def embed_mp3(music_file, meta_tags):
     """ Embed metadata to MP3 files. """
     # EasyID3 is fun to use ;)
+    # For supported easyid3 tags:
+    # https://github.com/quodlibet/mutagen/blob/master/mutagen/easyid3.py
+    # Check out somewhere at end of above linked file
     audiofile = EasyID3(music_file)
     audiofile['artist'] = meta_tags['artists'][0]['name']
     audiofile['albumartist'] = meta_tags['artists'][0]['name']
@@ -59,7 +62,8 @@ def embed_mp3(music_file, meta_tags):
     audiofile['arranger'] = meta_tags['artists'][0]['name']
     audiofile['performer'] = meta_tags['artists'][0]['name']
     audiofile['website'] = meta_tags['external_urls']['spotify']
-    audiofile['length'] = str(meta_tags['duration_ms'] / 1000)
+    audiofile['length'] = str(meta_tags['duration_ms'] / 1000.0)
+
     if meta_tags['publisher']:
         audiofile['encodedby'] = meta_tags['publisher']
     if meta_tags['genre']:
@@ -69,14 +73,27 @@ def embed_mp3(music_file, meta_tags):
     if meta_tags['external_ids']['isrc']:
         audiofile['isrc'] = meta_tags['external_ids']['isrc']
     audiofile.save(v2_version=3)
+
+    # For supported id3 tags:
+    # https://github.com/quodlibet/mutagen/blob/master/mutagen/id3/_frames.py
+    # Each class represents an id3 tag
     audiofile = ID3(music_file)
+    print(meta_tags['release_date'].split('-')[0])
+    year, *_ = meta_tags['release_date'].split('-')
+    audiofile['TORY'] = TORY(encoding=3, text=year)
+    audiofile['TYER'] = TYER(encoding=3, text=year)
+    audiofile['TPUB'] = TPUB(encoding=3, text=meta_tags['publisher'])
+    audiofile['COMM'] = COMM(encoding=3, text=meta_tags['external_urls']['spotify'])
+    if meta_tags['lyrics']:
+        audiofile['USLT'] = USLT(encoding=3, desc=u'Lyrics', text=meta_tags['lyrics'])
     try:
         albumart = urllib.request.urlopen(meta_tags['album']['images'][0]['url'])
-        audiofile["APIC"] = APIC(encoding=3, mime='image/jpeg', type=3,
+        audiofile['APIC'] = APIC(encoding=3, mime='image/jpeg', type=3,
                                  desc=u'Cover', data=albumart.read())
         albumart.close()
     except IndexError:
         pass
+
     audiofile.save(v2_version=3)
     return True
 
@@ -89,6 +106,7 @@ def embed_m4a(music_file, meta_tags):
              'artist'       : '\xa9ART',
              'date'         : '\xa9day',
              'title'        : '\xa9nam',
+             'year'         : '\xa9day',
              'originaldate' : 'purd',
              'comment'      : '\xa9cmt',
              'group'        : '\xa9grp',
@@ -100,7 +118,8 @@ def embed_m4a(music_file, meta_tags):
              'cpil'         : 'cpil',
              'albumart'     : 'covr',
              'copyright'    : 'cprt',
-            'tempo': 'tmpo'}
+             'tempo'        : 'tmpo',
+             'lyrics'       : '\xa9lyr' }
 
     audiofile = MP4(music_file)
     audiofile[tags['artist']] = meta_tags['artists'][0]['name']
@@ -111,11 +130,16 @@ def embed_m4a(music_file, meta_tags):
                                        meta_tags['total_tracks'])]
     audiofile[tags['disknumber']] = [(meta_tags['disc_number'], 0)]
     audiofile[tags['date']] = meta_tags['release_date']
+    year, *_ = meta_tags['release_date'].split('-')
+    audiofile[tags['year']] = year
     audiofile[tags['originaldate']] = meta_tags['release_date']
+    audiofile[tags['comment']] = meta_tags['external_urls']['spotify']
     if meta_tags['genre']:
         audiofile[tags['genre']] = meta_tags['genre']
     if meta_tags['copyright']:
         audiofile[tags['copyright']] = meta_tags['copyright']
+    if meta_tags['lyrics']:
+        audiofile[tags['lyrics']] = meta_tags['lyrics']
     try:
         albumart = urllib.request.urlopen(meta_tags['album']['images'][0]['url'])
         audiofile[tags['albumart']] = [MP4Cover(

requirements.txt

@@ -6,3 +6,4 @@ mutagen >= 1.37
 unicode-slugify >= 0.1.3
 titlecase >= 0.10.0
 logzero >= 1.3.1
+lyricwikia >= 0.1.8

spotdl.py

@@ -9,6 +9,7 @@ from titlecase import titlecase
 from slugify import slugify
 import spotipy
 import pafy
+import lyricwikia
 import urllib.request
 import os
 import sys
@@ -57,6 +58,19 @@ def generate_metadata(raw_song):
     meta_tags[u'publisher'] = album['label']
     meta_tags[u'total_tracks'] = album['tracks']['total']
 
+    log.debug('Fetching lyrics')
+
+    try:
+        meta_tags['lyrics'] = lyricwikia.get_lyrics(
+                        meta_tags['artists'][0]['name'],
+                        meta_tags['name'])
+    except lyricwikia.LyricsNotFound:
+        meta_tags['lyrics'] = None
+
+    # remove unused clutter when debug meta_tags
+    del meta_tags['available_markets']
+    del meta_tags['album']['available_markets']
+
     log.debug(pprint.pformat(meta_tags))
     return meta_tags
 
@@ -99,15 +113,15 @@ def generate_youtube_url(raw_song, meta_tags, tries_remaining=5):
     else:
         song = generate_songname(meta_tags)
         query['q'] = song
-    log.debug('Query: {0}'.format(query))
+    log.debug('query: {0}'.format(query))
 
     data = pafy.call_gdata('search', query)
-    query2 = {'part': 'contentDetails,snippet,statistics',
+    query_results = {'part': 'contentDetails,snippet,statistics',
               'maxResults': 50,
               'id': ','.join(i['id']['videoId'] for i in data['items'])}
-    log.debug('Query2: {0}'.format(query2))
+    log.debug('query_results: {0}'.format(query_results))
 
-    vdata = pafy.call_gdata('videos', query2)
+    vdata = pafy.call_gdata('videos', query_results)
 
     videos = []
     for x in vdata['items']:
@@ -122,8 +136,6 @@ def generate_youtube_url(raw_song, meta_tags, tries_remaining=5):
     if not videos:
         return None
 
-    log.debug(pprint.pformat(videos))
-
     if args.manual:
         log.info(song)
         log.info('0. Skip downloading this song.\n')
@@ -491,8 +503,8 @@ if __name__ == '__main__':
         elif args.username:
             feed_playlist(username=args.username)
 
-        # Actually we don't necessarily need this, but yeah...
-        # Explicit is better than implicit!
+        # actually we don't necessarily need this, but yeah...
+        # explicit is better than implicit!
         sys.exit(0)
 
     except KeyboardInterrupt as e: