Description

If one enters a simply URL[[BR]] to renders to a <a>URLBR</a>

Steps to reproduce

1. go to WikiSandBox and enter: http://openvpn.sourceforge.net[[BR]]

Examples

• http://example.com/[[BR]]

• http://example.com/[[BR

• http://example.com/[[]]

• http://example.com/[[

• http://example.com/[

• http://example.com/]]

• http://example.com/]

• http://example.com/{{{BR}}}

• http://example.com/{{{}}}

• http://example.com/{{{BR

• http://example.com/{{{

• http://example.com/}}}

• http://example.com/{{

• http://example.com/}}

Details

This Wiki.

Workaround

put a " " (space) between the URL and [[BR]]

Discussion

This is not necessarily a bug. Just a question of precedence. Why do you think that it qualifies as one?

This is a bug, but not because of the missing space, but because the closing ']]' are not part of the url. URLs should end only on whitespace. Happen also with other markup.

• How about tokens that are not permitted in URLs? Like umlauts etc. Maybe it should stop on them too.

If we stop on anything that not allowed in URL, then we can't have url like this: http://example.com/ניר_סופר . If you try this url, you will see that the browser will quote the url for you, and it appear as http://example.com/%D7%A0%D7%99%D7%A8_%D7%A1%D7%95%D7%A4%D7%A8 in the location box. The same thing will happen with any non ascii character. So if we want to support any url, we must stop only on whitespace. Checking this in the rfc would be a good idea.

• I don't think that any RFC talks about URL recognition in free form text files.

   1 punct_pattern = re.escape(u'''"\'}]|:,.)?!''')

These chars are not allowed at the end of an URL - which is an good idea in most cases. But they are allowed inside URLs. So the produces HTML is surprising but nevertheless correct.

Plan

• Priority:
• Assigned to:
• Status: not a Bug

CategoryMoinMoinNoBug