From 2686e3df7cd149b8bbb6b6c632b73c4bbe2366a5 Mon Sep 17 00:00:00 2001
From: Asad Dhamani <dhamaniasad@gmail.com>
Date: Sat, 4 Apr 2015 07:31:58 +0530
Subject: [PATCH] Switch to UnicodeDammit for encoding detection

---
 readability/encoding.py | 63 +++++++++++++----------------------------
 setup.py                |  3 +-
 2 files changed, 21 insertions(+), 45 deletions(-)

diff --git a/readability/encoding.py b/readability/encoding.py
index fb4761df..5df36eaa 100644
--- a/readability/encoding.py
+++ b/readability/encoding.py
@@ -1,48 +1,23 @@
-import re
-import chardet
+from bs4 import UnicodeDammit
 
 def get_encoding(page):
-    # Regex for XML and HTML Meta charset declaration
-    charset_re = re.compile(r'<meta.*?charset=["\']*(.+?)["\'>]', flags=re.I)
-    pragma_re = re.compile(r'<meta.*?content=["\']*;?charset=(.+?)["\'>]', flags=re.I)
-    xml_re = re.compile(r'^<\?xml.*?encoding=["\']*(.+?)["\'>]')
-
-    declared_encodings = (charset_re.findall(page) +
-            pragma_re.findall(page) +
-            xml_re.findall(page))
-
-    # Try any declared encodings
-    if len(declared_encodings) > 0:
-        for declared_encoding in declared_encodings:
-            try:
-                page.decode(custom_decode(declared_encoding))
-                return custom_decode(declared_encoding)
-            except UnicodeDecodeError:
-                pass
-
-    # Fallback to chardet if declared encodings fail
-    text = re.sub('</?[^>]*>\s*', ' ', page)
-    enc = 'utf-8'
-    if not text.strip() or len(text) < 10:
-        return enc # can't guess
-    res = chardet.detect(text)
-    enc = res['encoding'] or 'utf-8'
-    #print '->', enc, "%.2f" % res['confidence']
-    enc = custom_decode(enc)
+    # Pass in html to UnicodeDammit for encoding detection
+    page = UnicodeDammit(page)
+    enc = page.original_encoding
     return enc
 
-def custom_decode(encoding):
-    """Overrides encoding when charset declaration
-       or charset determination is a subset of a larger
-       charset.  Created because of issues with Chinese websites"""
-    encoding = encoding.lower()
-    alternates = {
-        'big5': 'big5hkscs',
-        'gb2312': 'gb18030',
-        'ascii': 'utf-8',
-        'MacCyrillic': 'cp1251',
-    }
-    if encoding in alternates:
-        return alternates[encoding]
-    else:
-        return encoding
+# def custom_decode(encoding):
+#     """Overrides encoding when charset declaration
+#        or charset determination is a subset of a larger
+#        charset.  Created because of issues with Chinese websites"""
+#     encoding = encoding.lower()
+#     alternates = {
+#         'big5': 'big5hkscs',
+#         'gb2312': 'gb18030',
+#         'ascii': 'utf-8',
+#         'MacCyrillic': 'cp1251',
+#     }
+#     if encoding in alternates:
+#         return alternates[encoding]
+#     else:
+#         return encoding
diff --git a/setup.py b/setup.py
index e7bb5884..bfa66983 100755
--- a/setup.py
+++ b/setup.py
@@ -24,7 +24,8 @@
     packages=['readability'],
     install_requires=[
         "chardet",
-        lxml_requirement
+        lxml_requirement,
+        "beautifulsoup4 >= 4.3.2"
         ],
     classifiers=[
         "Environment :: Web Environment",