Patrick, in my opinion software should be tolerant in cases where sometimes
you just want a result--for example, when providing print-on-demand via a
server, where customers have no chance to try to fix the source, or when
you are developing a review document that you know is not perfect, but you
want the programmers to review other content regardless of the validity.
Too-strict termination of process means that you cannot also collect
information on other potential errors until you fix this one issue.
Dealing with duplicate IDs is really a problem with any reuse architecture,
whether XInclude, DITA's conref or topicref, or even reuse based on content
management referencing methods. Therefore I conclude that FOP's particular
abend response is hyper-strict in a situation that is better served by just
a message. I credit the other formatters for dealing with the case in a
"Do What I Mean" manner.
Regards,
--
Don Day
Chair, OASIS DITA Technical Committee
IBM Lead DITA Architect
Email: dond at us.ibm.com
11501 Burnet Rd. MS9033E015, Austin TX 78758
Phone: +1 512-838-8550
T/L: 678-8550
"Where is the wisdom we have lost in knowledge?
Where is the knowledge we have lost in information?"
--T.S. Eliot
"Patrick Riffel"
<versorger at gmx.li
> To
Sent by: "xmleditor-support at xmlmind.com"
xmleditor-support <xmleditor-support at xmlmind.com>
-bounces at xmlmind. cc
com
Subject
Re: [XXE] (no subject)
08/23/2006 07:28
AM
> I'll post a RFE?/Bug report? to the FOP mailing list. (If you could do
> the same thing, it would be nice: two votes are better than one.)
I am no longer sure whom to blame. I would even go so far to say that the
fop behaviour is somewhat correct: The docbook stylesheets try to generate
a (hyperlinked) list of tables at the beginning of a document. If the doc
has two table elements with the same ID one on page X and the other on page
Y, where should the corresponding links point to -- the first or second
instance of the ID?
I tried a sample document (transformed with the docbook stylesheets 1.70.1)
with XEP 4.4 (the free edition). As you have already said, XEP does not
complain about the duplicate ID but the result is not correct (at least not
in my opinion): The list of tables correctly shows the table two times but
with identically page numbers.
To me it seems it more and more to be a conceptual problem in the usage of
xi:include along with IDs rather than a formatter issue.
puzzled
Patrick
-------- Original-Nachricht --------
Datum: Wed, 23 Aug 2006 11:00:24 +0200
Von: Hussein Shafie <hussein at xmlmind.com>
An: Patrick Riffel <versorger at gmx.li>
Betreff: Re: [XXE] (no subject)
> Patrick Riffel wrote:
> >
> >>Sure, we do that all the time here at XMLmind. The main difference
> >>between you and us is that we do not feel the need to double-check our
> >>documents using xmllint.
> >
> > Well, it's less a reason of double-checking than one of automation
> issues. I use XXE as an editor in a "classical" docbook toolchain: Part
of that
> is two resolve xincludes via xmllint, do the xsl transformation with
saxon
> and create the PDF with apache's fop. And while I could very well live
with
> a duplicate ID, fop throws an error and stops transformation.
>
> I understand. This is really a problem.
>
> * Saxon (or is it the XSLT style sheet?) just issues this warning:
> Warning: multiple "IDs" for constraint linkend: limitation_list.
>
> * Both FOP 0.20.5 and FOP 0.92beta throw a ValidationException and stop
> working.
>
> * RenderX XEP happily converts the XSL-FO to PDF.
>
> * Same with XMLmind FO Converter.
>
>
>
> > Any idea how to deal with this? BTW, the 3.3 feature to ignore such
> errors sounds promising, at least for the editing part ...
>
> Well, with FOP 0.92beta, using configuration parameter
> strict-validation=false or using command-line option -r should have
> forced FOP to ignore such benign validation error. But unfortunately I
> cannot get that to work.
>
> I'll post a RFE?/Bug report? to the FOP mailing list. (If you could do
> the same thing, it would be nice: two votes are better than one.)
--
"Feel free" ? 10 GB Mailbox, 100 FreeSMS/Monat ...
Jetzt GMX TopMail testen: http://www.gmx.net/de/go/topmail
--
XMLmind XML Editor Support List
xmleditor-support at xmlmind.com
http://www.xmlmind.com/mailman/listinfo/xmleditor-support
