Docs update

HISTORY.rst

@@ -3,8 +3,11 @@ Changelog
 
 0.6.3 (unreleased)
 ------------------
-- Word tokenization fix: Words that stem from a contraction will still have an apostrophe, e.g. ``"Let's" => ["Let", "'s"]``. This makes it easier to identify contractions.
+.. module:: text.taggers
+
+- Word tokenization fix: Words that stem from a contraction will still have an apostrophe, e.g. ``"Let's" => ["Let", "'s"]``.
 - Fix bug with comparing blobs to strings.
+- Add :class:`text.taggers.PerceptronTagger <PerceptronTagger>`, a fast and accurate POS tagger. Thanks `@syllog1sm <http://github.com/syllog1sm>`_.
 - Note for Python 3 users: You may need to update your corpora, since NLTK master has reorganized its corpus system. Just run ``curl https://raw.github.com/sloria/TextBlob/master/download_corpora.py | python`` again.
 - Add ``download_corpora_lite.py`` script for getting the minimum corpora requirements for TextBlob's basic features.
 

docs/install.rst

@@ -57,6 +57,23 @@ To get the latest development version of TextBlob, run
 
     $ pip install -U git+https://github.com/sloria/TextBlob.git@dev
 
+Getting Extra Models and Data
+-----------------------------
+
+.. module:: text.taggers
+
+Some features, such as the :class:`PerceptronTagger <PerceptronTagger>`, require data that is not available from the NLTK downloader. These data will be made available on the Github `release page`_ for TextBlob.
+
+To install a model or corpus:
+
+1. Download the file from the `release page`_.
+2. Unzip/untar the downloaded file.
+3. Place the uncompressed file in your TextBlob installation directory. To find out where this is, you can run ::
+
+    $ python -c "import text; print(text.__path__[0])"
+
+.. _release page: https://github.com/sloria/TextBlob/releases
+
 Python
 ++++++
 

text/_perceptron.py

@@ -56,6 +56,7 @@ def upd_feat(c, f, w, v):
             weights = self.weights.setdefault(f, {})
             upd_feat(truth, f, weights.get(truth, 0.0), 1.0)
             upd_feat(guess, f, weights.get(guess, 0.0), -1.0)
+        return None
 
     def average_weights(self):
         for feat, weights in self.weights.items():
@@ -68,14 +69,16 @@ def average_weights(self):
                 if averaged:
                     new_feat_weights[clas] = averaged
             self.weights[feat] = new_feat_weights
+        return None
 
     def save(self, path):
         '''Save the pickled model weights.'''
-        pickle.dump(dict(self.weights), open(path, 'w'))
+        return pickle.dump(dict(self.weights), open(path, 'w'))
 
     def load(self, path):
         '''Load the pickled model weights.'''
         self.weights = pickle.load(open(path))
+        return None
 
 
 def train(nr_iter, examples):

text/taggers.py

@@ -66,8 +66,14 @@ class PerceptronTagger(BaseTagger):
 
     '''Greedy Averaged Perceptron tagger, as implemented by Matthew Honnibal.
     Requires that ``trontagger.pickle`` exists in the text/ package directory.
+    For more information on how to get and install the pickled model, see
+    the install guide here:
 
-    See more info at this blog post: http://honnibal.wordpress.com/2013/09/11/a-good-part-of-speechpos-tagger-in-about-200-lines-of-python/
+        https:/textblob.readthedocs.org/en/latest/install.html
+
+    See more info about the Averaged Perceptron Tagger at this blog post:
+
+        http://honnibal.wordpress.com/2013/09/11/a-good-part-of-speechpos-tagger-in-about-200-lines-of-python/
 
     The tagger is about 96.8%% accurate.