Fix #1666: malformed binary data in SoundCheck

beets/mediafile.py

@@ -215,9 +215,9 @@ def _sc_decode(soundcheck):
     # SoundCheck tags consist of 10 numbers, each represented by 8
     # characters of ASCII hex preceded by a space.
     try:
-        soundcheck = soundcheck.replace(' ', '').decode('hex')
+        soundcheck = soundcheck.replace(b' ', b'').decode('hex')
         soundcheck = struct.unpack(b'!iiiiiiiiii', soundcheck)
-    except (struct.error, TypeError, UnicodeEncodeError):
+    except (struct.error, TypeError):
         # SoundCheck isn't in the format we expect, so return default
         # values.
         return 0.0, 0.0

docs/changelog.rst

@@ -56,6 +56,8 @@ Fixes:
 * :doc:`/plugins/duplicates`: Fix a crash when merging items. :bug:`1699`
 * :doc:`/plugins/smartplaylist`: More gracefully handle malformed queries and
   missing configuration.
+* Fix a crash with some files with unreadable iTunes SoundCheck metadata.
+  :bug:`1666`
 
 .. _Emby Server: http://emby.media
 

test/test_mediafile_edge.py

@@ -89,6 +89,13 @@ def test_only_magic_bytes_jpeg(self):
             beets.mediafile._image_mime_type(jpg_data),
             'image/jpeg')
 
+    def test_soundcheck_non_ascii(self):
+        # Make sure we don't crash when the iTunes SoundCheck field contains
+        # non-ASCII binary data.
+        f = beets.mediafile.MediaFile(os.path.join(_common.RSRC,
+                                                   'soundcheck-nonascii.m4a'))
+        self.assertEqual(f.rg_track_gain, 0.0)
+
 
 class InvalidValueToleranceTest(unittest.TestCase):
 
@@ -269,19 +276,19 @@ def test_round_trip(self):
         self.assertEqual(peak, 1.0)
 
     def test_decode_zero(self):
-        data = u' 80000000 80000000 00000000 00000000 00000000 00000000 ' \
-               u'00000000 00000000 00000000 00000000'
+        data = b' 80000000 80000000 00000000 00000000 00000000 00000000 ' \
+               b'00000000 00000000 00000000 00000000'
         gain, peak = beets.mediafile._sc_decode(data)
         self.assertEqual(gain, 0.0)
         self.assertEqual(peak, 0.0)
 
     def test_malformatted(self):
-        gain, peak = beets.mediafile._sc_decode(u'foo')
+        gain, peak = beets.mediafile._sc_decode(b'foo')
         self.assertEqual(gain, 0.0)
         self.assertEqual(peak, 0.0)
 
     def test_special_characters(self):
-        gain, peak = beets.mediafile._sc_decode(u'caf\xe9')
+        gain, peak = beets.mediafile._sc_decode(u'caf\xe9'.encode('utf8'))
         self.assertEqual(gain, 0.0)
         self.assertEqual(peak, 0.0)