Fix audio thumbnailing once and for all.

This change adds a Python 3-specific audio thumbnailer that side-steps the
bundled Python 2-only `audioprocessing` module. Instead of an audio spectrogram,
Python 3 users will get a static image.

This also allows me to remove my ineffective customisations to
`audioprocessing`, returning it to the upstream version as should always be the
case for vendored code in "extlib".

extlib/freesound/audioprocessing.py

@@ -43,32 +43,7 @@ def get_sound_type(input_filename):
 try:
     import scikits.audiolab as audiolab
 except ImportError:
-    print("WARNING: audiolab is not installed so wav2png will not work")
-
-    # Hack to prevent errors when uploading audio files. The issue is that
-    # scikits.audiolab does not support Python 3. By replacing it with a mock
-    # implementation here, we can accept audio files, but we won't get the nice
-    # waveform image.
-    import six
-    if six.PY3:
-        class MockSndfile(object):
-            def __init__(self, *args, **kwargs):
-                self.nframes = 0
-                self.channels = 1
-                self.samplerate = 44100
-
-            def read_frames(self, *args):
-                return []
-
-            def seek(self, *args):
-                return
-
-            def close(self):
-                return
-        import unittest.mock as mock
-        audiolab = mock.Mock()
-        audiolab.Sndfile = MockSndfile
-
+    print "WARNING: audiolab is not installed so wav2png will not work"
 import subprocess
 
 class AudioProcessingException(Exception):