file-based genre whitelist & cleanup

--HG--
rename : beetsplug/lastgenre.py => beetsplug/lastgenre/__init__.py

MANIFEST.in

@@ -8,3 +8,6 @@ include LICENSE NEWS README.rst
 # Resources for web plugin.
 recursive-include beetsplug/web/templates *
 recursive-include beetsplug/web/static *
+
+# And for the lastgenre plugin.
+include beetsplug/lastgenre/genres.txt

beetsplug/lastgenre/__init__.py

@@ -13,16 +13,30 @@
 # included in all copies or substantial portions of the Software.
 
 """Gets genres for imported music based on Last.fm tags.
+
+Uses a provided whitelist file to determine which tags are valid genres.
+The genre whitelist can be specified like so in .beetsconfig:
+
+    [lastgenre]
+    whitelist=/path/to/genres.txt
+
+The included (default) genre list was produced by scraping Wikipedia.
+The scraper script used is available here:
+https://gist.github.com/1241307
 """
+from __future__ import with_statement
 
 import logging
 import pylast
+import os
 
-from beets import plugins, ui
+from beets import plugins
+from beets import ui
+from beets.util import normpath
 
 log = logging.getLogger('beets')
 LASTFM = pylast.LastFMNetwork(api_key=plugins.LASTFM_KEY)
-WEIGHT_THRESH = 50
+DEFAULT_WHITELIST = os.path.join(os.path.dirname(__file__), 'genres.txt')
 
 def _tags_for(obj):
     """Given a pylast entity (album or track), returns a list of
@@ -44,30 +58,40 @@ def _tags_for(obj):
     return tags
 
 def _tags_to_genre(tags):
-    """Given a tag list, returns a genre. Returns the first tag that is present
-    in genres white list or None if no tag is suitable. 
+    """Given a tag list, returns a genre. Returns the first tag that is
+    present in the genre whitelist or None if no tag is suitable. 
     """
     if not tags:
         return None
-    elif not options['genres_whitelist']:
+    elif not options['whitelist']:
         return tags[0].title()
     
-    for tag in tags :
-        if tag.lower() in options['genres_whitelist'] :
+    for tag in tags:
+        if tag.lower() in options['whitelist']:
             return tag.title()
     
     return None
 
 options = {
-    'genres_whitelist': None,
+    'whitelist': None,
 }
 class LastGenrePlugin(plugins.BeetsPlugin):
     def configure(self, config):
-        genres_whitelist = ui.config_val(config, 'lastgenre',
-                                         'genres_whitelist', None)
-        if genres_whitelist :
-            genres_whitelist = genres_whitelist.lower().split(',')
-        options['genres_whitelist'] = genres_whitelist
+        wl_filename = ui.config_val(config, 'lastgenre', 'whitelist', None)
+        if not wl_filename:
+            # No filename specified. Instead, use the whitelist that's included
+            # with the plugin (inside the package).
+            wl_filename = DEFAULT_WHITELIST
+        wl_filename = normpath(wl_filename)
+
+        # Read the whitelist file.
+        whitelist = set()
+        with open(wl_filename) as f:
+            for line in f:
+                line = line.decode('utf8').strip().lower()
+                if line:
+                    whitelist.add(line)
+        options['whitelist'] = whitelist
         
 @LastGenrePlugin.listen('album_imported')
 def album_imported(lib, album):

docs/changelog.rst

@@ -5,7 +5,10 @@ Changelog
 -----------------------
 
 * The new :doc:`/plugins/lastgenre` automatically assigns genres to imported
-  albums and items based on Last.fm tags.
+  albums and items based on Last.fm tags and an internal whitelist. (Thanks to
+  `KraYmer`_.)
+
+.. _KraYmer: https://github.com/KraYmer
 
 1.0b10 (September 22, 2011)
 ---------------------------

docs/plugins/lastgenre.rst

@@ -21,10 +21,20 @@ your ``plugins`` line in :doc:`/reference/config`, like so::
     [beets]
     plugins: lastgenre
 
-For the time being, the genre-selection algorithm is pretty dumb: the most
-popular tag is treated as the genre. This could be enhanced by using a "white
-list" of known genre names. (This would be a great project for someone looking
-to contribute to the beets project!)
+The plugin chooses genres based on a *whitelist*, meaning that only certain tags
+can be considered genres. This way, tags like "my favorite music" or "seen live"
+won't be considered genres. The plugin ships with a fairly extensive internal
+whitelist, but you can set your own in the config file using the ``whitelist``
+configuration value::
+
+    [lastgenre]
+    whitelist: /path/to/genres.txt
+
+The genre list file should contain one genre per line. Blank lines are ignored.
+
+For the curious, the default genre list is generated by a `script that scrapes
+Wikipedia`_.
 
 .. _pip: http://www.pip-installer.org/
 .. _pylast: http://code.google.com/p/pylast/
+.. _script that scrapes Wikipedia: https://gist.github.com/1241307

setup.py

@@ -31,7 +31,7 @@ setup(name='beets',
       platforms='ALL',
       long_description=_read('README.rst'),
       test_suite='test.testall.suite',
-      include_package_data=True, # Install web plugin resources.
+      include_package_data=True, # Install plugin resources.
 
       packages=[
           'beets',
@@ -41,6 +41,7 @@ setup(name='beets',
           'beetsplug',
           'beetsplug.bpd',
           'beetsplug.web',
+          'beetsplug.lastgenre',
       ],
       namespace_packages=['beetsplug'],
       entry_points={