[email protected]
[Top] [All Lists]

Bug#451267: marked as forwarded (dblatex: tries to compile incorrect doc

Subject: Bug#451267: marked as forwarded dblatex: tries to compile incorrect document
From: Debian Bug Tracking System
Date: Fri, 16 Nov 2007 22:21:03 +0000
Your message dated Fri, 16 Nov 2007 23:33:33 +0100
with message-id <[email protected]>
has caused the Debian Bug report #451267,
regarding dblatex: tries to compile incorrect document
to be marked as having been forwarded to the upstream software
author(s) .

(NB: If you are a system administrator and have no idea what I am
talking about this indicates a serious mail system misconfiguration
somewhere.  Please contact me immediately.)

Debian bug tracking system administrator
(administrator, Debian Bugs database)

--- Begin Message ---
Subject: Re: Bug#451267: dblatex: tries to compile incorrect document
From: "benoit.guillon"
Date: Fri, 16 Nov 2007 23:33:33 +0100
Hi Andreas,

On Fri, 16 Nov 2007 17:30:22 +0100, Andreas Hoenen <[email protected]> wrote:
In this specific case, the compilation fails, and that's expected.
However I'd like to see something about the real reason, which is
malformed XML, not something about a missing \item.

Thus it's a usability issue.  And it's not restricted to dblatex, but to
the XSLT processor [1].  I can imagine two approaches:

1) Force a validation before XSLT transformation.  This abandons the
   best effort approach, but is rather straight forward.

2) On transformation errors validate the document afterwards.  If
   validation fails, dump this as a possible error cause.

The problem for dblatex in general is that it is aimed at various
platforms, thus adding another dependency is likely to cause trouble.
However the Debian installation could patch the general product to add
such a validation.

Thus let's ask upstream:

Benoit, what do you think about this feature request?  Any objections
against a Debian specific extension?  Or do you prefer a general
solution, maybe in a way that validation will only occur if xmllint is
I would tend to think that XML validation is out of the scope of dblatex.
At least, a systematic validation before compiling is not an option (time
consuming, option to add to disable the new feature, etc.). The idea about
validating on error is an interesting alternative but I'd like to avoid
embedding unnecessary processing in dblatex.
This said, if you think this feature is interesting for debian packaged
dblatex, why not and feel free to patch (if you need some hints, don't
hesitate to ask). Just to know: how does FOP compilation behave on wrong

--- End Message ---
<Prev in Thread] Current Thread [Next in Thread>