autotag: Can now compute the distance for incomplete albums

If the user has some songs from a specific album, but not all of them,
the real solution is immediately discarded. This commit is the first of a
series that will implement support for these incomplete albums.

The point of this patch is to make sure missing commits are taken into
account when calculating the distance between an album and its canonical
data.

Note that in order not to break API compatibility, the album_distance
call for the plugins receives a purged version of both the items and the
album info, resulting in some potential accuracy if the plugin bases
itself on the index of a track in album_info.tracks.

beets/autotag/match.py

@@ -17,6 +17,7 @@ releases and tracks.
 """
 import logging
 import re
+import copy
 from munkres import Munkres
 from unidecode import unidecode
 
@@ -31,6 +32,8 @@ ARTIST_WEIGHT = 3.0
 ALBUM_WEIGHT = 3.0
 # The weight of the entire distance calculated for a given track.
 TRACK_WEIGHT = 1.0
+# The weight of a missing track.
+MISSING_WEIGHT = 0.3
 # These distances are components of the track distance (that is, they
 # compete against each other but not ARTIST_WEIGHT and ALBUM_WEIGHT;
 # the overall TRACK_WEIGHT does that).
@@ -161,7 +164,7 @@ def current_metadata(items):
     likelies = {}
     consensus = {}
     for key in keys:
-        values = [getattr(item, key) for item in items]
+        values = [getattr(item, key) for item in items if item]
         likelies[key], freq = plurality(values)
         consensus[key] = (freq == len(values))
     return likelies['artist'], likelies['album'], consensus['artist']
@@ -269,12 +272,25 @@ def distance(items, album_info):
     
     # Track distances.
     for i, (item, track_info) in enumerate(zip(items, album_info.tracks)):
-        dist += track_distance(item, track_info, i+1, album_info.va) * \
-                TRACK_WEIGHT
-        dist_max += TRACK_WEIGHT
+        if item:
+            dist += track_distance(item, track_info, i+1, album_info.va) * \
+                    TRACK_WEIGHT
+            dist_max += TRACK_WEIGHT
+        else:
+            dist += MISSING_WEIGHT
+            dist_max += MISSING_WEIGHT
 
     # Plugin distances.
-    plugin_d, plugin_dm = plugins.album_distance(items, album_info)
+    # In order not to break compatibility, send purged lists
+    purged_items, purged_tracks = [], []
+    for i, t in zip(items, album_info.tracks):
+        if i:
+            purged_items.append(i)
+            purged_tracks.append(t)
+    purged_album_info = copy.copy(album_info)
+    purged_album_info.tracks = purged_tracks
+
+    plugin_d, plugin_dm = plugins.album_distance(purged_items, purged_album_info)
     dist += plugin_d
     dist_max += plugin_dm
 

test/test_autotag.py

@@ -94,6 +94,21 @@ class AlbumDistanceTest(unittest.TestCase):
         )
         self.assertEqual(match.distance(items, info), 0)
 
+    def test_incomplete_album(self):
+        items = []
+        items.append(self.item('one', 1))
+        items.append(self.item('three', 3))
+        info = AlbumInfo(
+            artist = 'some artist',
+            album = 'some album',
+            tracks = self.trackinfo(),
+            va = False,
+            album_id = None, artist_id = None,
+        )
+        self.assertNotEqual(match.distance(items, info), 0)
+        # Make sure the distance is not too great
+        self.assertTrue(match.distance(items, info) < 0.2)
+
     def test_global_artists_differ(self):
         items = []
         items.append(self.item('one', 1))