Merge branch 'master' of ssh://github.com/honnibal/spaCy into develop

.appveyor.yml

@@ -8,16 +8,24 @@ environment:
   matrix:
 
     # Python 2.7.10 is the latest version and is not pre-installed.
-
     - PYTHON: "C:\\Python27.10-x64"
       PYTHON_VERSION: "2.7.10"
       PYTHON_ARCH: "64"
 
+    - PYTHON: "C:\\Python27.10-x32"
+      PYTHON_VERSION: "2.7.10"
+      PYTHON_ARCH: "32"
+
     # The lastest Python 3.4.
     - PYTHON: "C:\\Python34-x64" 
       PYTHON_VERSION: "3.4.x" # currently 3.4.3 
       PYTHON_ARCH: "64" 
 
+    #- PYTHON: "C:\\Python34-x32" 
+    #  PYTHON_VERSION: "3.4.x" # currently 3.4.3 
+    #  PYTHON_ARCH: "32" 
+
+
 install:
   # Install Python (from the official .msi of http://python.org) and pip when
   # not already installed.
@@ -30,10 +38,11 @@ install:
   - "SET PYTHONPATH=%CD%;%PYTHONPATH%"
 
   # Filesystem root
-  # - ps: "ls \"C:/\""
+  #- ps: "ls \"C:/\""
+  #- SET
 
   # Installed SDKs
-  # - ps: "ls \"C:/Program Files/Microsoft SDKs/Windows\""
+  #- ps: "ls \"C:/Program Files/Microsoft SDKs/Windows\""
 
   # Checking stdint.h
   #- ps: "ls \"C:/projects/spacy/include/\""

README.md

@@ -1,3 +1,6 @@
+
+<img src="https://ci.appveyor.com/api/projects/status/aoe3dtkep36rdaqf?svg=true" />
+
 spaCy: Industrial-strength NLP
 ==============================
 
@@ -49,3 +52,6 @@ Difficult to support:
 
 * PyPy 2.7
 * PyPy 3.4
+
+
+

spacy/en/download.py

@@ -1,5 +1,4 @@
 from __future__ import print_function
-from os import path
 import sys
 import os
 import tarfile
@@ -15,45 +14,44 @@ AWS_STORE = 'https://s3-us-west-1.amazonaws.com/media.spacynlp.com'
 
 ALL_DATA_DIR_URL = '%s/en_data_all-%s.tgz' % (AWS_STORE, VERSION)
 
-DEST_DIR = path.join(path.dirname(path.abspath(__file__)), 'data')
+DEST_DIR = os.path.dirname(os.path.abspath(__file__))
 
 
-def download_file(url, dest_dir):
-    return uget.download(url, dest_dir, console=sys.stdout)
+def download_file(url, download_path):
+    return uget.download(url, download_path, console=sys.stdout)
 
 
-def install_data(url, dest_dir):
-    filename = download_file(url, dest_dir)
-    t = tarfile.open(filename)
-    t.extractall(dest_dir)
+def install_data(url, extract_path, download_path):
+    try:
+        os.makedirs(extract_path)
+    except FileExistsError:
+        pass
 
-
-def install_parser_model(url, dest_dir):
-    filename = download_file(url, dest_dir)
-    t = tarfile.open(filename, mode=":gz")
-    t.extractall(dest_dir)
-
-
-def install_dep_vectors(url, dest_dir):
-    download_file(url, dest_dir)
+    tmp = download_file(url, download_path)
+    assert tmp == download_path
+    t = tarfile.open(download_path)
+    t.extractall(extract_path)
 
 
 @plac.annotations(
     force=("Force overwrite", "flag", "f", bool),
 )
 def main(data_size='all', force=False):
-    if data_size == 'all':
-        data_url = ALL_DATA_DIR_URL
-    elif data_size == 'small':
-        data_url = SM_DATA_DIR_URL
+    filename = ALL_DATA_DIR_URL.rsplit('/', 1)[1]
+    download_path = os.path.join(DEST_DIR, filename)
+    data_path = os.path.join(DEST_DIR, 'data')
 
-    if force and path.exists(DEST_DIR):
-        shutil.rmtree(DEST_DIR)
+    if force and os.path.exists(download_path):
+        os.unlink(download_path)
 
-    if not os.path.exists(DEST_DIR):
-        os.makedirs(DEST_DIR)
+    if force and os.path.exists(data_path):
+        shutil.rmtree(data_path)
 
-    install_data(data_url, DEST_DIR)
+    if os.path.exists(data_path):
+        print('data already installed at %s, overwrite with --force' % DEST_DIR)
+        sys.exit(1)
+
+    install_data(ALL_DATA_DIR_URL, DEST_DIR, download_path)
 
 
 if __name__ == '__main__':

spacy/lemmatizer.py

@@ -19,7 +19,7 @@ class Lemmatizer(object):
             index[pos] = read_index(path.join(data_dir, 'wordnet', 'index.%s' % pos))
             exc[pos] = read_exc(path.join(data_dir, 'wordnet', '%s.exc' % pos))
         if path.exists(path.join(data_dir, 'vocab', 'lemma_rules.json')):
-            rules = json.load(open(path.join(data_dir, 'vocab', 'lemma_rules.json')))
+            rules = json.load(codecs.open(path.join(data_dir, 'vocab', 'lemma_rules.json'), encoding='utf_8'))
         else:
             rules = {}
         return cls(index, exc, rules)

spacy/tokens/doc.pyx

@@ -120,6 +120,9 @@ cdef class Doc:
     def __str__(self):
         return u''.join([t.string for t in self])
 
+    def __repr__(self):
+        return u''.join([t.string for t in self])
+
     def similarity(self, other):
         if self.vector_norm == 0 or other.vector_norm == 0:
             return 0.0

spacy/tokens/spans.pyx

@@ -46,6 +46,12 @@ cdef class Span:
             return 0
         return self.end - self.start
 
+    def __repr__(self):
+        text = self.text_with_ws
+        if self[-1].whitespace_:
+            text = text[:-1]
+        return text
+
     def __getitem__(self, object i):
         if isinstance(i, slice):
             start, end = normalize_slice(len(self), i.start, i.stop, i.step)

spacy/tokens/token.pyx

@@ -43,6 +43,9 @@ cdef class Token:
     def __str__(self):
         return self.string
 
+    def __repr__(self):
+        return self.string
+
     cpdef bint check_flag(self, attr_id_t flag_id) except -1:
         return Lexeme.c_check_flag(self.c.lex, flag_id)