W3C home > Mailing lists > Public > xproc-dev@w3.org > August 2014

No-XML inputs

From: Wendell Piez <wapiez@wendellpiez.com>
Date: Fri, 22 Aug 2014 17:08:28 -0400
Message-ID: <CAAO_-xyWw244LysFBU7Czw+SVkh84uO-T0zehvj-hHMeWhQABA@mail.gmail.com>
To: XProc Dev <xproc-dev@w3.org>
Hi,

So somewhere between xmlcalabash-1.0.3-94 and 1.0.18-95, my XProcs
stopped working. I now get an error about a p:document failing to
load. From the command line, the system even hangs. On breaking it I
can see in the tb:

SEVERE: It is a dynamic error if the resource referenced by a
p:document element does not exist, cannot be accessed, or is not a
well-formed XML document.

Admittedly these are very rude creatures, these XProcs. They call in
non-XML input by accepting a runtime option (a string) they can hand
to Saxon, which is then invited to use unparsed-text() to read them.

The XProc itself never sees anything but XML - the front end of its
input stream is given as a literal in the XProc:

<p:input port="source">
  <p:inline><document><!-- dummy document</document</p:inline>
</p:input>
<p:with-param name="file" select="$fileOption"/>

Then Saxon does unparsed-text($file) to get at the data.

It's still working in oXygen, btw -- even though it has 1.0.18-95.
Whatever is in the way preventing Saxon from getting the file in
Calabash (and then parsing it), isn't intruding in oXygen.

So I am wondering what is wrong and how to fix it. (In the meantime I
am not throwing away my old Calabash. Being able to parse and wf-check
LMNL syntax from the cl is too important to me to lose.)

In general, what is the recommended practice currently when the
primary input is not XML? I think I was playing hooky the day that was
discussed.

Thanks for your thoughts --
Wendell


-- 
Wendell Piez | http://www.wendellpiez.com
XML | XSLT | electronic publishing
Eat Your Vegetables
_____oo_________o_o___ooooo____ooooooo_^
Received on Sunday, 24 August 2014 08:36:58 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:03:12 UTC