Thanks for looking into it on your end.
I suppose I should properly blame xsltproc, because xmlto is just
a script to call it (and a parser beforehand). For completeness,
included is some screen output:
$ xmlto -vv html temp.xml
Format script: /usr/share/xmlto/format/docbook/html
Convert to HTML (with chunks)
xmllint >/dev/null --xinclude --postvalid
/home/delanuez/temp/temp.xml
Stylesheet: http://docbook.sourceforge.net/release/xsl/current/html/chunk.xsl
xsltproc --nonet --xinclude \
-o /tmp/xmlto.oH2424/temp.proc
\
http://docbook.sourceforge.net/release/xsl/current/html/chunk.xsl
\
/home/delanuez/temp/temp.xml
Thanks again. I suppose that's
that. I can deal. :-)
David de la Nuez
(914)945-1167 (tie line 862-1167)
COIN-OR www.coin-or.org
"Bob Stayton"
<>
09/17/2004 01:29 PM
To
David De la nuez/Watson/IBM@IBMUS
cc
<>
Subject
Re: [docbook-apps] Question
links when making a qandaset into HTML
If you get "index.html" in
hrefs, then xmlto is not working correctly. The html/docbook.xsl
stylesheet doesn't know the name of the output filename, and so cannot
insert a filename into the references. When I process a qandaset
document with html/docbook.xsl, all of the hrefs start with "#".
The bit about the title is because
the stylesheets assume the root element of a document has a title. That
usually is not a bad assumption. 8^)
Bob Stayton
Sagehill Enterprises
DocBook Consulting
----- Original Message -----
From: David
De la nuez
To: Bob
Stayton
Cc:
Sent: Friday, September 17, 2004 10:11 AM
Subject: Re: [docbook-apps] Question links when
making a qandaset into HTML
Thanks for the reply, Bob.
I get the same results using the "html-nochunks" argument to
xmlto but I have to include a title in the qandaset or it won't work. The
verbose option shows me that it is using the current docbook.xsl sheet
(sorry I am not terribly versed in the fine points of stylesheets).
In any case, the reason that the "index.html" is a problem is
that when I include the .proc file in an HTML file which isn't named index.html...
the links aren't correct. It strikes me as odd default behavior.
It seems to a novice such as I am that for a qandaset the answers
are always going to be in the same chunk (or file) as the questions so
a simple "#id......" would work. In any case, that seems
more likely than the answers being in a file named "index.html".
It's not a big problem because I manually include the .proc into my HTML
file, so one more step is a trivial amount of extra work. I'm sure
there's a smarter way to do this, but this was the first that occurred
to me (though I think someone just asked a question recently about how
to do something similar in a more clever way?).
David de la Nuez
(914)945-1167 (tie line 862-1167)
COIN-OR www.coin-or.org
"Bob Stayton"
<>
09/17/2004 12:52 PM
To
<>,
David De la nuez/Watson/IBM@IBMUS
cc
Subject
Re: [docbook-apps] Question
links when making a qandaset into HTML
Hi David,
With the chunking stylesheets, all hrefs start with the filename.html.
Otherwise the stylesheet would have to compute if the target of the cross
reference is in the same chunk as the link text.
I don't understand why the "index.html" in the href is a problem.
Did you try using the nonchunking stylesheet instead?
Bob Stayton
Sagehill Enterprises
DocBook Consulting
----- Original Message -----
From: David
De la nuez
To:
Sent: Friday, September 17, 2004 7:42 AM
Subject: [docbook-apps] Question links when making a qandaset into
HTML
When I convert to HTML a DB XML document with qandaset as the root element,
the question links at the top of the resulting .proc page are all of the
form "index.html#idxxxxxx". It seems to me that the "index.html"
should not be there. The only reason I care is because to reproduce
on a web site in a quick and dirty way an FAQ I have in a DB book, this
is what I do. Is this a bug or a design decision I don't understand?
FYI, to the best of my knowledge I am using what the xmlto script uses
in cygwin, namely xsltproc with the current chunk.xsl stylesheet.
David de la Nuez
(914)945-1167 (tie line 862-1167)
COIN-OR www.coin-or.org
