Re: [xsl] Encoding problem or what else?

[Geert Josten <Geert.Josten@xxxxxxxxxxx>]

> Geert,
> this is interesting to know.
> What do you mean by "patch"?
> Do you mean perhaps that I should write something that strips out the 3 
> bytes from the beginning of the file?

Jup, I meant writing (Java?) code. But not in the sense of writing a separate app that bites ;) the 
three bytes off the head of the document, but merely adjusting the reading process in existing code. 
Provided you have access to it. When using the XSL parser in a larger framework (Cocoon perhaps), 
you can often do this fairly easy. When using the XSL parser from the command-line, typically not.

> I think that the easiest solution is to ask the people who deliver this 
> file to switch to ISO-8859-1 as there is no real need to use unicode for 
> these files, I mean, there is not going to be any text containing exotic 
> characters in there.

Jup, that is in line with my second suggestion. But perhaps they can use a different creation tool. 
This problem is most heared when people are editing XML documents with a text editor.

> I am bound to use this xsl processor for the simple reason that it's the 
> best of the bunch from a performance standpoint (thanks Micheal Kay!).
> I've been struggling for days with Altova XSLT 2005 engine and Oracle's 
> internal processor and it was a nightmare.
> I had a file of 32Mb xml file that took *hours* to be processed with 
> these two processors until I tried out saxon that cruched it in less 
> than one minute!

Nice..

> So, as you can easily guess, I am not going to willingly dump Saxon just 
> for those three funny bytes.

No, you shouldn't. But perhaps someone knows a way to configure Saxon such that it uses a different 
XML parser front end?

> Hey Micheal, what do you think about this?
> Is there any hope that xerces will "consume" this utf-8 marker in the 
> near future?

Cheers..