[fix] utf8 content highlighting

Browse Source
This commit is contained in:
asciimoo 2013-11-19 11:27:17 +01:00
parent d0427d9bae
commit 3755f437e4
1 changed files with 7 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
searx/engines/__init__.py
View File

@ -112,18 +112,19 @@ def highlight_content(content, query):
if content.find('<') != -1: if content.find('<') != -1:
return content return content
query = query.decode('utf-8')
if content.lower().find(query.lower()) > -1: if content.lower().find(query.lower()) > -1:
query_regex = '({0})'.format(re.escape(query)) query_regex = u'({0})'.format(re.escape(query))
content = re.sub(query_regex, '<b>\\1</b>', content, flags=re.I) content = re.sub(query_regex, '<b>\\1</b>', content, flags=re.I | re.U)
else: else:
regex_parts = [] regex_parts = []
for chunk in query.split(): for chunk in query.split():
if len(chunk) == 1: if len(chunk) == 1:
regex_parts.append('\W+{0}\W+'.format(re.escape(chunk))) regex_parts.append(u'\W+{0}\W+'.format(re.escape(chunk)))
else: else:
regex_parts.append('{0}'.format(re.escape(chunk))) regex_parts.append(u'{0}'.format(re.escape(chunk)))
query_regex = '({0})'.format('|'.join(regex_parts)) query_regex = u'({0})'.format('|'.join(regex_parts))
content = re.sub(query_regex, '<b>\\1</b>', content, flags=re.I) content = re.sub(query_regex, '<b>\\1</b>', content, flags=re.I | re.U)
return content return content