OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

docbook-apps message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [docbook-apps] Multiple processing document with exsl:node-setand xsltproc


On Mon, 11 Aug 2003 13:18:46 +0200
Jirka Kosek <jirka@kosek.cz> wrote:

> Daniel Veillard wrote:
> 
> >   Well do the maintainers expect to fix the stylesheets or
> >   feel fine
> > keeping code based un undefined behaviour ? This sounds
> > insane to just copy full subtree of the documents to then
> > look for keys in a copy of the given set ... Can't profiling
> > be done without copying to a new subtree ???
> 
> Profiling doesn't copy full subtree it copies just subset of it
> which is determined by profiling conditions. Doing profiling
> without creating new subtree is almost impossible -- it would
> require very complex change to almost every XPath expression in
> the stylesheets.
> 
> I always know that single-pass profiling operates in a gray
> fuzzy zone. Doing two stage profiling using temporary document
> is completely safe and works in every processor. The profiling
> transformation is very easy and runs very fast so there is no
> big impact on performance compared to the single-pass
> profiling.

In some cases saving temporary result to file is not good. It's
require rewriting Makefiles to using temporary files as input
instead source XML.

<skipped/>

-- 
Regards, Vyt
mailto:  vyt@vzljot.ru
JID:     vyt@vzljot.ru


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]