takegue · August 29, 2015 14:14 · Jan 24, 2015 · Jan 24, 2015 · Jan 24, 2015
diff --git a/2015-01-25-005613.py b/2015-01-25-005613.py
@@ -4,14 +4,17 @@
 import subprocess
 import itertools as itt
 
+
 class MeCab():
 
-    def __init__(self,opts=[]):
+    def __init__(self, opts=[]):
         self.opts = opts
         self._process = subprocess.Popen(
             list(itt.chain(['mecab'], opts)),
             stdin=subprocess.PIPE,
-            stdout=subprocess.PIPE)
+            stdout=subprocess.PIPE,
+            universal_newlines =True,
+        )
 
         if '-Owakati' in opts or 'wakati' in opts:
             self.parse = self.wakati_parse
@@ -30,7 +33,8 @@ def default_parse(self, iterable):
             self._process.stdin.write(line+'\n')
             while True:
                 output = self._process.stdout.readline().strip()
-                if output.startswith('EOS'): break
+                if output.startswith('EOS'):
+                    break
                 buff.append(Morph.from_mecab_format(output))
             yield buff
             buff = []
@@ -53,10 +57,11 @@ def from_mecab_format(cls, line):
         morph.pos = attrs[0]
         return morph
 
+
 def test_mecab():
     parser = MeCab(['-Owakati'])
-    morphs =  next(parser.parse(['すもももももももものうち'] * 10))
-    assert morphs == ['すもも','も','もも','も','もも','の','うち']
+    morphs = next(parser.parse(['すもももももももものうち']))
+    assert morphs == ['すもも', 'も', 'もも', 'も', 'もも', 'の', 'うち']
 
 
 if __name__ == '__main__':

diff --git a/2015-01-25-005613.py b/2015-01-25-005613.py
@@ -0,0 +1,63 @@
+#!/usr/bin/env python
+# -*- coding:utf-8 -*-
+
+import subprocess
+import itertools as itt
+
+class MeCab():
+
+    def __init__(self,opts=[]):
+        self.opts = opts
+        self._process = subprocess.Popen(
+            list(itt.chain(['mecab'], opts)),
+            stdin=subprocess.PIPE,
+            stdout=subprocess.PIPE)
+
+        if '-Owakati' in opts or 'wakati' in opts:
+            self.parse = self.wakati_parse
+        else:
+            self.parse = self.default_parse
+
+    def wakati_parse(self, iterable):
+        for line in iterable:
+            self._process.stdin.write(line+'\n')
+            output = self._process.stdout.readline()
+            yield output.strip().split()
+
+    def default_parse(self, iterable):
+        buff = []
+        for line in iterable:
+            self._process.stdin.write(line+'\n')
+            while True:
+                output = self._process.stdout.readline().strip()
+                if output.startswith('EOS'): break
+                buff.append(Morph.from_mecab_format(output))
+            yield buff
+            buff = []
+
+
+class Morph():
+
+    def __init__(self):
+        self.surface = ''
+        self.attrs = []
+
+    @classmethod
+    def from_mecab_format(cls, line):
+        surface, attrs = line.split()
+        attrs = attrs.split(',')
+
+        morph = Morph()
+        morph.surface = surface
+        morph.attrs = attrs
+        morph.pos = attrs[0]
+        return morph
+
+def test_mecab():
+    parser = MeCab(['-Owakati'])
+    morphs =  next(parser.parse(['すもももももももものうち'] * 10))
+    assert morphs == ['すもも','も','もも','も','もも','の','うち']
+
+
+if __name__ == '__main__':
+    test_mecab()