From f0f6bb16244c97ca07645367b66d2454fcd7cfc3 Mon Sep 17 00:00:00 2001 From: Oleg Broytman Date: Tue, 4 Mar 2008 09:42:31 +0000 Subject: [PATCH] Replace ISO-8859-2 to the default encoding. git-svn-id: file:///home/phd/archive/SVN/bookmarks_db/trunk@194 fdd5c36f-1aea-0310-aeeb-c58d7e2b6c23 --- Robots/parse_html_beautifulsoup.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Robots/parse_html_beautifulsoup.py b/Robots/parse_html_beautifulsoup.py index 162e7a3..f76f058 100644 --- a/Robots/parse_html_beautifulsoup.py +++ b/Robots/parse_html_beautifulsoup.py @@ -77,7 +77,7 @@ def parse_html(filename, charset=None, log=None): head = html # Some sites put TITLE in HTML without HEAD _charset = root.originalEncoding - if _charset == "windows-1252": # Replace default + if _charset in ("ISO-8859-2", "windows-1252"): # Replace default _charset = DEFAULT_CHARSET try: -- 2.39.2