RE: document() runs out of memory

OK, thanks for trying.

> Can someone confirm that the XSL segment I included in my initial post
will read in each file, one at a time, process it and then read in the next
without creating a huge union of all of the file contents?

Well, it does not create a union.  But, Xalan stuffs the documents in a
hashtable right now, and keeps it around until the processing is done.  It
really ought to keep them in a limited LRU pool.  If you are using the
Xerces DOM liaison, this will be more expensive because it will create a
full DOM each time.

It's open source, so if you want to work with us to put in a LRU pool, we
would be glad to help.  You could also install your own version of the
Document function.

-scott




                                                                                                                           
                    Brad Sommerfeld                                                                                        
                    <bsommerfeld@xxxxxxx>        To:     "'Scott_Boag@xxxxxxxxx'" <Scott_Boag@xxxxxxxxx>                   
                    Sent by:                     cc:     xsl-list@xxxxxxxxxxxxxxxx                                         
                    owner-xsl-list@mulber        Subject:     RE: document() runs out of memory                            
                    rytech.com                                                                                             
                                                                                                                           
                                                                                                                           
                    03/21/00 05:21 PM                                                                                      
                    Please respond to                                                                                      
                    xsl-list                                                                                               
                                                                                                                           
                                                                                                                           



I just tried the Xalan 1.0.0 version with the same results.  I'm not
convinced the issue is the processor as much as the XSL that I have
constructed.


Can someone confirm that the XSL segment I included in my initial post will
read in each file, one at a time, process it and then read in the next
without creating a huge union of all of the file contents?


Also, I tried the code suggested by Myriam and it wouldn't parse with a
reference to a variable within the variable declaration.


> Xalan keeps an internal cache of the documents, so it shouldn't be
> recreating a source tree or parsing each time.  I suspect you
> are running
> into a variable bug that has since been fixed.  We just released Xalan
> 1.0.0, so I suggest you try that.
>








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