- From: Dominique Hazael-Massieux <dom@w3.org>
- Date: Wed, 11 Jul 2007 11:57:20 +0200
- To: Rafe Harwood <rafe@rafe.me.uk>
- Cc: public-qa-dev@w3.org
Hi Rafe, Le mercredi 11 juillet 2007 à 02:18 +0100, Rafe Harwood a écrit : > When viewing the results page, the "URI of the HTML document to extract > semantics from:" text entry box, uses an encoded string instead of the > entered string from the previous page. This means that if that information > is entered (mistakenly clicking the submit button or another piece added to > the end of the current string), the system returns a java exception error. > > It may be a good idea (if possible) to keep that as an un-encoded string for > ease of manipulation and removing the chance of getting the java exception > error. Indeed, this would be much preferable! Unfortunately, this bug is a consequence of the tool being based on an XSLT style sheet, with the limitations that this language has - indeed, there is no readily available function in xpath 2 to convert a URL-encoded string into a decoded one; if I had time, I would look into writing such a function and adding it to the extractor, but I fear I won't be able to do so very soon. Should you or anyone on the list feel like writing such a function, I would gladly integrate it in the tool. Regards, Dom
Received on Wednesday, 11 July 2007 09:57:33 UTC