Last modified: 2005-04-13 17:11:10 UTC
Basically MediaWiki generates an <A HREF...> link for the ISBN leading to a "special" page. This link is retained as part of the image alt-text (created based on the description). Of course, having a link inside ALT attribute is not healthy. Some HTML code spills out of the tag onto the page. You can see the problem at the link provided with this bug entry (I saw it in Opera 8, but I guess most browsers should work the same here).
We should not use such offensive languages here.
It is easy not to include ISBNs in image captions.
The problem is that Parser::magicISBN() is called after Linker::makeThumbLinkObj has been run on the text, which means that the magicISBN function will expand on any ISBN number it recognizes, including those inside XHTML tags like alt="" or title="".
(In reply to comment #3) > The problem is that Parser::magicISBN() is called after Linker::makeThumbLinkObj > has been run on the text, which means that the magicISBN function will expand on > any ISBN number it recognizes, including those inside XHTML tags like alt="" or > title="". In that case, it should be possible to do things the other way round, since the image code should strip the HTML tags back out of the alt/title attributes. Kind of ugly, but I think it might work.
*** This bug has been marked as a duplicate of 1887 ***
