Merge branch 'master' of ssh://github.com/spacy-io/spaCy

requirements.txt

@@ -5,7 +5,7 @@ cymem>=1.30,<1.32
 preshed>=0.46.1,<0.47.0
 thinc>=5.0.0,<5.1.0
 murmurhash>=0.26,<0.27
-plac
+plac<0.9.3
 six
 ujson
 cloudpickle

spacy/download.py

@@ -9,15 +9,16 @@ from sputnik.package_list import (PackageNotFoundException,
 from . import about
 
 
-def download(lang, force=False):
+def download(lang, force=False, fail_on_exist=True):
     if force:
         sputnik.purge(about.__title__, about.__version__)
 
     try:
         sputnik.package(about.__title__, about.__version__, about.__models__[lang])
-        print("Model already installed. Please run 'python -m "
-              "spacy.%s.download --force' to reinstall." % lang, file=sys.stderr)
-        sys.exit(1)
+        if fail_on_exist:
+            print("Model already installed. Please run 'python -m "
+                  "spacy.%s.download --force' to reinstall." % lang, file=sys.stderr)
+            sys.exit(0)
     except (PackageNotFoundException, CompatiblePackageNotFoundException):
         pass
 

website/blog/sense2vec-with-spacy.jade

@@ -37,7 +37,7 @@ p Here's how the current pre-processing function looks, at the time of writing.
                         # Merge them into single tokens
                         ent.merge(ent.root.tag_, ent.text, ent.label_)
             token_strings = []
-            for token in tokens:
+            for token in doc:
                 text = token.text.replace(' ', '_')
                 tag = token.ent_type_ or token.pos_
                 token_strings.append('%s|%s' % (text, tag))

website/docs/_quickstart-examples.jade

@@ -46,13 +46,13 @@
         assert token.shape_ == 'Xxxxx'
         for lexeme in nlp.vocab:
             if lexeme.is_alpha:
-                lexeme.shape_ = 'W'
+                lexeme.shape_ = u'W'
             elif lexeme.is_digit:
-                lexeme.shape_ = 'D'
+                lexeme.shape_ = u'D'
             elif lexeme.is_punct:
-                lexeme.shape_ = 'P'
+                lexeme.shape_ = u'P'
             else:
-                lexeme.shape_ = 'M'
+                lexeme.shape_ = u'M'
         assert token.shape_ == 'W'
 
     +h3('examples-numpy-arrays') Export to numpy arrays