http://blah.net/bla/faq.php#5 <- that's php DASH FIVE

edit
woops.. that works indeed.
I was using urls like:
0000005" rel="noopener,nofollow">http://blah.net/bla/faq/0000005 <- thats faq SLASH DASH FIVE
which is causing problems.

sorry for the confusion.

edited on: 08-04-04 18:39

OK, so the problem is that bug references are being expanded inside URLs.

This has been reported in other fields. For exampe, HTML tags can show up in headings (ref 0004126).

edited on: 08-06-04 11:42

This problem appears with other characters too. For example &0004126. Are there any plans to escape those bug link tags or note tags other than changing the config file?

I think using # char for hotlinking bug numbers was an unfortunate choice.

That said, please note that using plain numbers for anchors (like in <a name="5"> is an error according to w3c standards. see:

http://www.w3.org/TR/html4/types.html#type-cdata

So the current matching regex's are allowing hyperlinks containing foo.html#1 ( foo dot html hash 1), but not allowing a hyperlink containing /0000001 (slash hash 1)

As mentioned by giallu in 0004271:0017848, id/name tags are not allowed to start with a digit and must match the pattern [A-Za-z][A-Za-z0-9:_.-]* as per html 4.01 standard [1]. In later versions of the standard, it is recommended to stick to this pattern for backwards-compatibility [2], even though technically a wider range of characters is permitted [3]

Based on the above, I'm resolving this as "won't fix".

For those who would want to alter MantisBT to match syzop's requirements, the regex in string_process_bug_link() function can be altered to exclude additional chars, e.g. the '/' in the example below:

diff --git a/core/string_api.php b/core/string_api.php

index 67b6f9e..ad1f751 100644

--- a/core/string_api.php

+++ b/core/string_api.php

@@ -368,7 +368,7 @@ function string_process_bug_link( $p_string, $p_include_anchor = true, $p_detail

}

}


$p_string = preg_replace_callback( '/(^|[^\w&])' . preg_quote( $t_tag, '/' ) . '(\d+)\b/', $string_process_bug_link_callback[$p_include_anchor][$p_detail_info][$p_fqdn], $p_string );
$p_string = preg_replace_callback( '/(^|[^\w&\/])' . preg_quote( $t_tag, '/' ) . '(\d+)\b/', $string_process_bug_link_callback[$p_include_anchor][$p_detail_info][$p_fqdn], $p_string );

return $p_string;

}

[1] http://www.w3.org/TR/html4/types.html#type-id
[2] http://www.w3.org/TR/html/#C_8
[3] http://www.w3.org/TR/xml/#NT-Nmtoken