Merge pull request #2206 from Kraymer/issue2205

Issue2205: search for lyrics using part of song title preceding ':'

beetsplug/lyrics.py

@@ -25,8 +25,8 @@ import re
 import requests
 import unicodedata
 import warnings
-from six.moves import urllib
 import six
+from six.moves import urllib
 
 try:
     from bs4 import SoupStrainer, BeautifulSoup
@@ -144,29 +144,32 @@ def search_pairs(item):
     The method also tries to split multiple titles separated with `/`.
     """
 
+    def generate_alternatives(string, patterns):
+        """Generate string alternatives by extracting first matching group for
+           each given pattern."""
+        alternatives = [string]
+        for pattern in patterns:
+            match = re.search(pattern, string, re.IGNORECASE)
+            if match:
+                alternatives.append(match.group(1))
+        return alternatives
+
     title, artist = item.title, item.artist
-    titles = [title]
-    artists = [artist]
 
-    # Remove any featuring artists from the artists name
-    pattern = r"(.*?) {0}".format(plugins.feat_tokens())
-    match = re.search(pattern, artist, re.IGNORECASE)
-    if match:
-        artists.append(match.group(1))
+    patterns = [
+        # Remove any featuring artists from the artists name
+        r"(.*?) {0}".format(plugins.feat_tokens())]
+    artists = generate_alternatives(artist, patterns)
 
-    # Remove a parenthesized suffix from a title string. Common
-    # examples include (live), (remix), and (acoustic).
-    pattern = r"(.+?)\s+[(].*[)]$"
-    match = re.search(pattern, title, re.IGNORECASE)
-    if match:
-        titles.append(match.group(1))
-
-    # Remove any featuring artists from the title
-    pattern = r"(.*?) {0}".format(plugins.feat_tokens(for_artist=False))
-    for title in titles[:]:
-        match = re.search(pattern, title, re.IGNORECASE)
-        if match:
-            titles.append(match.group(1))
+    patterns = [
+        # Remove a parenthesized suffix from a title string. Common
+        # examples include (live), (remix), and (acoustic).
+        r"(.+?)\s+[(].*[)]$",
+        # Remove any featuring artists from the title
+        r"(.*?) {0}".format(plugins.feat_tokens(for_artist=False)),
+        # Remove part of title after colon ':' for songs with subtitles
+        r"(.+?)\s*:.*"]
+    titles = generate_alternatives(title, patterns)
 
     # Check for a dual song (e.g. Pink Floyd - Speak to Me / Breathe)
     # and each of them.

docs/changelog.rst

@@ -60,6 +60,8 @@ And there are a few bug fixes too:
   This is fixed. :bug:`2168`
 * :doc:`/plugins/embyupdate`: Fixes authentication header problem that caused
   a problem that it was not possible to get tokens from the Emby API.
+* :doc:`/plugins/lyrics`: Search for lyrics using the title part preceding the 
+  colon character. :bug:`2206`
 
 The last release, 1.3.19, also erroneously reported its version as "1.3.18"
 when you typed ``beet version``. This has been corrected.

test/test_lyrics.py

@@ -84,6 +84,10 @@ class LyricsPluginTest(unittest.TestCase):
         self.assertIn(('Alice', ['song']),
                       lyrics.search_pairs(item))
 
+        item = Item(artist='Alice and Bob', title='song')
+        self.assertEqual(('Alice and Bob', ['song']),
+                         list(lyrics.search_pairs(item))[0])
+
     def test_search_pairs_multi_titles(self):
         item = Item(title='1 / 2', artist='A')
         self.assertIn(('A', ['1 / 2']), lyrics.search_pairs(item))
@@ -118,6 +122,10 @@ class LyricsPluginTest(unittest.TestCase):
         self.assertNotIn(('A', ['Song']), lyrics.search_pairs(item))
         self.assertIn(('A', ['Song and B']), lyrics.search_pairs(item))
 
+        item = Item(title='Song: B', artist='A')
+        self.assertIn(('A', ['Song']), lyrics.search_pairs(item))
+        self.assertIn(('A', ['Song: B']), lyrics.search_pairs(item))
+
     def test_remove_credits(self):
         self.assertEqual(
             lyrics.remove_credits("""It's close to midnight