Bump version to 0.4; fix some changes to analyze/english

carkov/__init__.py

@@ -4,4 +4,4 @@
 # This is free software, see the included LICENSE for terms and conditions.
 #
 
-version = '0.2.0'
+version = '0.4.0'

carkov/analyze/english.py

@@ -4,16 +4,8 @@
 # This is free software, see the included LICENSE for terms and conditions.
 #
 
-import nltk
-
 from .abstract import AbstractAnalyzer
-
-
-try:
-    nltk.sent_tokenize("foo bar")
-except LookupError:
-    nltk.download('punkt')
-
+from .utils import english_sentence_split, english_sentence_tokenize
 
 class English(AbstractAnalyzer):
     def __init__(self, order, filters=None):
@@ -25,8 +17,8 @@ class English(AbstractAnalyzer):
         chunks = corpus.split('\n\n')
         ret = []
         for chunk in chunks:
-            ret = ret + nltk.sent_tokenize(chunk)
+            ret = ret + english_sentence_split(chunk)
         return ret
 
     def tokenize_segment(self, segment):
-        return list(nltk.word_tokenize(segment))
+        return list(english_sentence_tokenize(segment))

setup.cfg

@@ -1,6 +1,6 @@
 [metadata]
 name = carkov
-version = 0.2.0
+version = 0.4.0
 description = A markov chainer library
 author = Aldercone Studio
 author_email = alderconestudio@gmail.com