Re: coping with huge xml-saxon

> So what exactly is the memory limit for XSLT transformations? I know
> that Michale Kay in his book mentions that documents larger than 1 or 2
> MB cannot be processed by an XSLT processor, because, as you say, the
> processor loads them into memory.

Did he really say that?  That would surprise me, because that limit seems
rather low.  I have transformed 500MB documents with Xalan-C, and once
transformed a 1GB document.  That wasn't pretty, and there was a great deal
of swapping going on, but the transformation did complete.

> It seem that if you have 256 MB of memory, you ought to be able to
> handle a document of around 50 MB. (I think the processor actually
> requires something like 4 times the size of the document--or was it 10?)

There are many factors to consider.  Markup-heavy documents might take up
less space than content-heavy documents.  But really, it's very
implementation-specific.  I think jd-xslt has an option to page parts of
the document to disk, which would certainly help with large documents.
Xalan-C's default implementation of the source tree keeps the entire
document in memory but tries to be as compact as possible.  However,
someone could write an different implementation which keeps the majority of
the document in a database, etc.

Dave


 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list