You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@xalan.apache.org by bu...@apache.org on 2001/11/08 17:27:52 UTC
DO NOT REPLY [Bug 4735] -
Infinite loop processing large file
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://nagoya.apache.org/bugzilla/show_bug.cgi?id=4735>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND
INSERTED IN THE BUG DATABASE.
http://nagoya.apache.org/bugzilla/show_bug.cgi?id=4735
Infinite loop processing large file
keshlam@us.ibm.com changed:
What |Removed |Added
----------------------------------------------------------------------------
AssignedTo|xalan-dev@xml.apache.org |keshlam@us.ibm.com
------- Additional Comments From keshlam@us.ibm.com 2001-11-08 08:27 -------
One can quibble about whether this is a bug per se or simply a design limitation
of this implementation... but I'll accept it as a bug since we're looking for
ways to improve it.
Xalan-J does currently have a maxiumum size limit on input documents; we can run
out of slots in the DTM table, and I believe the failure mechanism could indeed
be seen as a wrap back to the root of the document. We should be able to make
that failure more diagnostically recognizable; I'll investigate that.
The best workaround I can offer for now is to subdivide the problem -- generate
a number of smaller documents rather than one huge one, and use document() to
access them. That will reduce the load on each DTM model. However, the same
issue means we can run out of available DTMs... so if you've got a particularly
large dataset, it may indeed be impossible to get it all into a single Xalan
execution and you may need to style separate chunks, then merge them via another
tool.
We do have some ideas which might address this, either by changing the DTM
handle representation (EXTREMELY painful, and we hope to avoid it) or by
introducing some source-tree pruning (which would be a good thing from many
points of view, but wouldn't cover all possible stylesheets.) A pruning
prototype is in progress but is not yet ready to be checked in, and at this time
our best guess is that it won't be included in reference releases of Xalan
before version 2.4.