Genres is a python library that analyzes text with musical context (such as reviews) in order to determine musical genres
How it works
Genres is based on a list of genres and tags (database), those are then compared against supplied text using regexp and points for ranking are calculated. Found genres are compared agains its main category to avoid mismatches.
The api is simple.
>>> import genres >>> r = genres.find("Pink Floyd is a rock band) >>> r ['rock']
It detects multiple genres, as long as they are related to the same category
>>> import genres >>> genres.find("Acid jazz, an electronic masterpiece.") ['acid jazz', 'jazz']
In this example the two occurences of electronic and techno triumps rock.
>>> import genres >>> genres.find("Electronic music with a techno vibe. Different from rock") ['techno', 'electronic']
The database is a simple list of words, separated by newline and structured like this:
|Metal,0.8||Sub category, Percentage of full rank (10*0.8=8)|
|-Pink Floyd||Tag associated to category rock|
|Categories are sparated with newline|
Genres are distributed with a database that can be found under
genres/data.txt and the genre structure is based on Allmusic genre categorisation.
It is possible to supply your own database:
import genres db_obj = genres.db.Db("./example.txt") finder_obj = genres.finder.Finder(db_obj)
Genres can easily be installed through pip.
$ pip install genres
This library include tests, just run
Want to contribute? Awesome. Just send a pull request.
Genres is released under the MIT License.