X-Git-Url: https://git.phdru.name/?a=blobdiff_plain;f=parse_html%2Fbkmk_ph_beautifulsoup4.py;h=1095ebce8d510c22df2c9768baed3c27777e1909;hb=48440e20170112c8a036b0c66c1cbe067ed1d87e;hp=aad3f8fb9092bdba2daec856a93f4309f9714795;hpb=11632b7623b2b2e01995f013bc6d8ba01c20cf74;p=bookmarks_db.git

diff --git a/parse_html/bkmk_ph_beautifulsoup4.py b/parse_html/bkmk_ph_beautifulsoup4.py
index aad3f8f..1095ebc 100644
--- a/parse_html/bkmk_ph_beautifulsoup4.py
+++ b/parse_html/bkmk_ph_beautifulsoup4.py
@@ -5,17 +5,20 @@ This file is a part of Bookmarks database and Internet robot.
 """
 
 __author__ = "Oleg Broytman <phd@phdru.name>"
-__copyright__ = "Copyright (C) 2017 PhiloSoft Design"
+__copyright__ = "Copyright (C) 2017-2023 PhiloSoft Design"
 __license__ = "GNU GPL"
 
 __all__ = ['parse_html']
 
 
-import re
 from bs4 import BeautifulSoup
+
 from .bkmk_ph_util import HTMLParser
+from compat import string_type
+
+universal_charset = "utf-8"
+DEFAULT_CHARSET = "cp1251"  # Stupid default for Russian Cyrillic
 
-DEFAULT_CHARSET = "cp1251" # Stupid default for Russian Cyrillic
 
 def _parse_html(html_text, charset):
     try:
@@ -23,6 +26,7 @@ def _parse_html(html_text, charset):
     except TypeError:
         return None
 
+
 def parse_html(html_text, charset=None, log=None):
     root = _parse_html(html_text, charset)
     if root is None:
@@ -35,7 +39,7 @@ def parse_html(html_text, charset=None, log=None):
 
     head = html.head
     if head is None:
-        head = html # Some sites put TITLE in HTML without HEAD
+        head = html  # Some sites put TITLE in HTML without HEAD
 
     title = head.title
     if (title is None) and (html is not head):
@@ -52,8 +56,8 @@ def parse_html(html_text, charset=None, log=None):
         else:
             parts = []
             for part in title:
-                if not isinstance(part, basestring):
-                    part = unicode(part)
+                if not isinstance(part, string_type):
+                    part = part.decode()
                 parts.append(part.strip())
             title = ''.join(parts)
 
@@ -62,10 +66,11 @@ def parse_html(html_text, charset=None, log=None):
         try:
             meta_content = meta.get("content")
             if meta_content:
-                __charset = meta_content.lower().split('charset=')[1].split(';')[0]
+                __charset = meta_content.lower().split('charset=')[1].\
+                    split(';')[0]
             else:
                 __charset = False
-        except IndexError: # No charset in the META Content-Type
+        except IndexError:  # No charset in the META Content-Type
             meta_charset = False
         else:
             meta_charset = _charset = __charset
@@ -79,8 +84,12 @@ def parse_html(html_text, charset=None, log=None):
             if meta_content:
                 meta_charset = _charset = meta_content.lower()
 
-    if title and (_charset or meta_charset):
-        title = title.encode(_charset or meta_charset)
+    #if title and (_charset or meta_charset):
+    #    try:
+    #        title = title.encode(_charset or meta_charset)
+    #    except LookupError:
+    #        title = title.encode(universal_charset)
+    #        _charset = universal_charset
 
     meta = head.find(_find_refresh, recursive=False)
     if meta:
@@ -98,17 +107,22 @@ def parse_html(html_text, charset=None, log=None):
         return None
     return HTMLParser(_charset, meta_charset, title, refresh, icon)
 
+
 def _find_contenttype(Tag):
     return (Tag.name == "meta") and \
        (Tag.get_attribute_list("http-equiv", '')[0].lower() == "content-type")
 
+
 def _find_charset(Tag):
     return (Tag.name == "meta") and Tag.get("charset", '')
 
+
 def _find_refresh(Tag):
     return (Tag.name == "meta") and \
        (Tag.get_attribute_list("http-equiv", '')[0].lower() == "refresh")
 
+
 def _find_icon(Tag):
     return (Tag.name == "link") and \
-       (Tag.get_attribute_list("rel", '')[0].lower() in ('icon', 'shortcut icon'))
+       (Tag.get_attribute_list("rel", '')[0].lower()
+        in ('icon', 'shortcut icon'))