bs4.1.0にcchardetを適用

bs4.1.0_cchardet_patch.diff

--- E:/bs4.orig/dammit.py	Wed Jun 13 13:52:56 2012
+++ E:/bs4/dammit.py	Wed Jun 20 17:30:47 2012
@@ -13,15 +13,11 @@
 import warnings
 
 # Autodetects character encodings. Very useful.
-# Download from http://chardet.feedparser.org/
-#  or 'apt-get install python-chardet'
-#  or 'easy_install chardet'
+# Download from https://github.com/PyYoshi/cChardet
 try:
-    import chardet
-    #import chardet.constants
-    #chardet.constants._debug = 1
+    import cchardet
 except ImportError:
-    chardet = None
+    cchardet = None
 
 # Available from http://cjkpython.i18n.org/.
 try:
@@ -207,8 +203,8 @@
                         break
 
         # If no luck and we have auto-detection library, try that:
-        if not u and chardet and not isinstance(self.markup, unicode):
-            u = self._convert_from(chardet.detect(self.markup)['encoding'])
+        if not u and cchardet and not isinstance(self.markup, unicode):
+            u = self._convert_from(cchardet.detect(self.markup))
 
         # As a last resort, try utf-8 and windows-1252:
         if not u:

tests.py

# coding:utf8

# nosetests --nocapture --with-profile --profile-stats-file nose.prof tests.py
# hotshot2dot nose.prof| dot -Tpng -o profile.png

import time
from bs4 import BeautifulSoup

class TestBS4():
    def test_speed(self):
        html = file(r"testdata1.htm").read()
        start = time.time()
        s = BeautifulSoup(html)
        end = time.time()
        print end - start