You are viewing a plain text version of this content. The canonical link for it is here.
Posted to fop-dev@xmlgraphics.apache.org by Rob Heittman <ro...@cluestream.com> on 2001/09/28 16:09:39 UTC
XSLTInputHandler constructor for InputSources?
Would it cause problems to provide an alternate constructor for
XSLTInputHandler ... ?
XSLTInputHandler(InputSource xmlsource, InputSource xsltsource)
in addition to:
XSLTInputHandler(File xmlfile, File xsltfile)
I am working with XML which is streamed directly out of a repository, not
a file, and could get rid of a bunch of lines of redundant code in my app
if I could use FOP's XSLTInputHandler instead.
If this makes sense, let me know and I'll be happy to provide a patch.
- Rob
--
Rob Heittman
cluestream ventures | effective information management
research, development, analysis and consulting
http://www.cluestream.com
---------------------------------------------------------------------
To unsubscribe, e-mail: fop-dev-unsubscribe@xml.apache.org
For additional commands, email: fop-dev-help@xml.apache.org
Re: XSLTInputHandler constructor for InputSources?
Posted by Matt Savino <ma...@synergizethis.com>.
I am in the exact same situation as far as my xml source. I've currently
been using a DOM XSLT transform and sending that result to FOP.
I built the latest CVS so I could try SAX (since the SAX input is broken
in .20.1). But as you'll see below, the times seem about the same. I
also don't know what to do about the logger errors in the latest cvs.
And finally I really like the way .20.1 shows you where the page breaks
are as it reports it's progress back to the program. Is that gone for
good? I think keeping the page numbers in rows rather than streaming
down the screen makes the console output easier to read for developers.
By the way I did a little rough benchmarking on FOP embedded in a
servlet. (I'm running on a Weblogic development box - Wintel - P-III -
933). I tried three different methods: a DOM source, a SAX stream from
an XSLT transfrom, and the XSLTInputHandler (I serialzed the xml source
out to a file). The XSLT transform is pretty simple as the xml-data is
already well structured.
Interestingly all three took between 50-51 seconds to render a ~5000
line report into a 300k, 200 page PDF document (there were a lot of page
breaks). I'm thinking that's not too bad. Especially since in our case I
don't think the end user will be able to tell the difference between
render time and download time.
Do these numbers sound about right?
-Matt
Rob Heittman wrote:
>
> Would it cause problems to provide an alternate constructor for
> XSLTInputHandler ... ?
>
> XSLTInputHandler(InputSource xmlsource, InputSource xsltsource)
>
> in addition to:
>
> XSLTInputHandler(File xmlfile, File xsltfile)
>
> I am working with XML which is streamed directly out of a repository, not
> a file, and could get rid of a bunch of lines of redundant code in my app
> if I could use FOP's XSLTInputHandler instead.
>
> If this makes sense, let me know and I'll be happy to provide a patch.
>
> - Rob
>
> --
>
> Rob Heittman
>
> cluestream ventures | effective information management
> research, development, analysis and consulting
> http://www.cluestream.com
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: fop-dev-unsubscribe@xml.apache.org
> For additional commands, email: fop-dev-help@xml.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: fop-dev-unsubscribe@xml.apache.org
For additional commands, email: fop-dev-help@xml.apache.org