Thread: [Refdb-users] xhtml references don't start on new line

Status: Beta

Brought to you by: mhoenicka

refdb-users

[Refdb-users] xhtml references don't start on new line

From: David N. <dav...@bi...> - 2004-03-10 12:59:54

Hi all,

In the latest CVS version of refdb, I have found that xhtml output does 
not start each reference on a new line.  Rather, they are on the same 
line, one running on after the other.

html and pdf output render correctly.

Regards,
David.

[Refdb-users] xhtml references don't start on new line

From: Markus H. <mar...@mh...> - 2004-03-10 22:14:33

David Nebauer writes:
 > Hi all,
 > 
 > In the latest CVS version of refdb, I have found that xhtml output does 
 > not start each reference on a new line.  Rather, they are on the same 
 > line, one running on after the other.
 > 

I'm afraid this may be a display problem of your browser. The output
looks just fine over here (Firebird 0.7 on FreeBSD).

 > html and pdf output render correctly.
 > 

Not quite. While comparing the output with xhtml, I've found two typos
in the html stylesheet. One of them screwed up bold vs italic
rendering which made the output look quite funny. This is fixed now in
CVS.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [Refdb-users] xhtml references don't start on new line

From: David N. <dav...@bi...> - 2004-03-11 11:26:15

Hi Markus,

>I'm afraid this may be a display problem of your browser. The output
>looks just fine over here (Firebird 0.7 on FreeBSD).
> =20
>

Problem is visible on both:

Firefox version 0.8-3 (Mozilla 1.6, Copyright (c) 2003 mozilla.org,=20
build 2004022702)

and

Firebird version 0.7-7 (Mozilla 1.5, Copyright (c) 2003 mozilla.org=20
<developer build>),

running on Debian Sarge/testing (these are the current versions in testin=
g).

I don't have Netscape installed.

KDE 's browser, Konqueror, wouldn't render the file.  The KHTML=20
rendering engine threw a parsing error on the Revision History table,=20
objecting to the <tr> element following the <table> element.  Perhaps it=20
wanted <tr> inside the <th> element?

If you trick lynx into rendering XHTML as HTML (see=20
<http://www.kolpackov.net/projects/LynxAndXHTML.xhtml>), it renders=20
correctly, but I'm not sure of the significance of this.

Here is the source xml document:
.........................................................................=
.................................................................
<?xml version=3D"1.0"?>
<!DOCTYPE article PUBLIC "-//OASIS//DTD DocBook XML V4.2//EN"=20
"http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd" [
<!ENTITY bibliography SYSTEM "reftoc.bib.xml">
]>
<article>
    <title>TOC Reference List tester</title>
    <articleinfo>
        <author>
            <surname>Nebauer</surname>
            <firstname>David</firstname>
            <affiliation>
                <address>
                    <email>dav...@bi...</email>
                </address>
            </affiliation>
        </author>
        <revhistory>
            <revision>
                <revnumber>1</revnumber>
                <date>2004-03-03</date>
                <authorinitials>DN</authorinitials>
                <revremark>Create document</revremark>
            </revision>
        </revhistory>
    </articleinfo>
    <!-- <abstract>
        <para>Demonstration of the &ldquo;missing TOC entry&rdquo;=20
phenomenon.</para>
    </abstract> -->
    <section id=3D"spurious-quotes">
        <title>Some quotes</title>
        <para>In order to interest refdb in my document, I have to make=20
some quotes, such as <citation role=3D"REFDB">Savitch2001</citation> and=20
<citation role=3D"REFDB">Lamport1994</citation>, for it to work on.  That=
=20
should suffice.</para>
    </section>
    <section id=3D"problem-statement">
        <title>Statement of the problem</title>
        <para>Some weeks ago I decided to plunge headlong into DocBook=20
XML.  With no prior experience I immersed myself in XSLT processors, FO=20
processors, XML, the DocBook specifications, catalogs (argh!) and,=20
finally, refdb.  As I proceeded I had a test document which steadily=20
grew in complexity.  When I reached the point of being able to ouptut=20
pdf, I noticed something strange about the Table of Contents.</para>
        <para>What I noticed was that there was an entry at the correct=20
place.  It gave page number and leader dots.  What was absent, however,=20
were the words &ldquo;Reference List&rdquo; at the start of the entry. =20
I notified <emphasis>refdb-users</emphasis> and continued.  My test=20
document grew increasingly more complex as I learned more.  Then I saw=20
an amazing thing.  The entry had corrected itself!  Yes, there was a=20
pristine Table of Contents entry right where it should have been all=20
along.</para>
        <para>My current working theory is that it wil be absent in=20
simple documents.  This doument is a test of that theory.  So, here goes=20
&hellip;</para>
    </section>
    &bibliography;
</article>
.........................................................................=
..............................................................

Here are the references to which it refers:

.........................................................................=
..............................................................
$ refdbc -C "getref -t risx :ID:<3"
<?xml version=3D"1.0" encoding=3D"ISO-8859-1"?>
<!DOCTYPE ris PUBLIC "-//Markus Hoenicka//DTD Ris V1.0.2//EN"=20
"http://refdb.sourceforge.net/dtd/risx-1.0.2/risx.dtd" [
<!ENTITY lt "&#38;#60;">
<!ENTITY gt "&#62;">
<!ENTITY amp "&#38;#38;">
]>
<ris>



  <entry type=3D"BOOK" id=3D"1" citekey=3D"Lamport1994">
    <publication>
      <title type=3D"gen">LaTeX: A Document Preparation System</title>
      <author>
        <lastname>Lamport</lastname>
        <firstname>Leslie</firstname>
      </author>
      <pubinfo>
        <pubdate type=3D"primary">
          <date><year>1994</year><month>00</month><day>00</day></date>
        </pubdate>
        <city>Massachusetts</city>
        <publisher>Addison-Wesley</publisher>
      </pubinfo>
    </publication>
    <libinfo user=3D"david">
      <reprint status=3D"NOTINFILE"/>
    </libinfo>
  </entry>


  <entry type=3D"BOOK" id=3D"2" citekey=3D"Savitch2001">
    <publication>
      <title type=3D"gen">Java. An Introduction to Computer Science and=20
Programming</title>
      <author>
        <lastname>Savitch</lastname>
        <firstname>Walter</firstname>
      </author>
      <pubinfo>
        <pubdate type=3D"primary">
          <date><year>2001</year><month>00</month><day>00</day></date>
        </pubdate>
        <city>New Jersey</city>
        <publisher>Prentice Hall</publisher>
      </pubinfo>
    </publication>
    <libinfo user=3D"david">
      <reprint status=3D"NOTINFILE"/>
    </libinfo>
  </entry>
</ris>
2 reference(s) retrieved

.........................................................................=
..............................................................

Here is the transformation command: refdbxml -s=20
Eur.J.Pharmacol.xhtml.xsl -t xhtml reftoc.xml

Note: XSLT processor =3D xsltproc.

Here is the resulting xhtml file:
.........................................................................=
..............................................................
$ cat reftoc.xhtml
<?xml version=3D"1.0" encoding=3D"UTF-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"=20
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns=3D"http://www.w3.org/1999/xhtml"><head><meta=20
http-equiv=3D"Content-Type" content=3D"text/html; charset=3DUTF-8"=20
/><title>TOC Reference List tester</title><meta name=3D"generator"=20
content=3D"DocBook XSL Stylesheets V1.64.1" /></head><body><div=20
class=3D"article" lang=3D"en" xml:lang=3D"en"><div=20
class=3D"titlepage"><div><div><h2 class=3D"title"><a id=3D"id2403246"></a=
>TOC=20
Reference List tester</h2></div><div><div class=3D"author"><h3=20
class=3D"author"><span class=3D"firstname">David</span> <span=20
class=3D"surname">Nebauer</span></h3><div class=3D"affiliation"><div=20
class=3D"address"><p><br />
                                        <tt class=3D"email">&lt;<a=20
href=3D"mailto:dav...@bi...">dav...@bi...</a>&gt;=
</tt><br=20
/>
                                </p></div></div></div></div><div><div=20
class=3D"revhistory"><table border=3D"1" width=3D"100%" summary=3D"Revisi=
on=20
history"><tr><th align=3D"left" valign=3D"top" colspan=3D"3"><b>Revision=20
History</b></th></tr><tr><td align=3D"left">Revision 1</td><td=20
align=3D"left">2004-03-03</td><td align=3D"left">DN</td></tr><tr><td=20
align=3D"left" colspan=3D"3">Create=20
document</td></tr></table></div></div></div><div></div><hr /></div><div=20
class=3D"toc"><p><b>Table of Contents</b></p><dl><dt><span=20
class=3D"section"><a href=3D"#spurious-quotes">Some=20
quotes</a></span></dt><dt><span class=3D"section"><a=20
href=3D"#problem-statement">Statement of the=20
problem</a></span></dt><dt><span class=3D"bibliography"><a=20
href=3D"#id2502984">Reference List</a></span></dt></dl></div><div=20
class=3D"section" lang=3D"en" xml:lang=3D"en"><div=20
class=3D"titlepage"><div><div><h2 class=3D"title" style=3D"clear: both"><=
a=20
id=3D"spurious-quotes"></a>Some=20
quotes</h2></div></div><div></div></div><p>In order to interest refdb in=20
my document, I have to make some quotes, such as <a xmlns=3D""=20
href=3D"#IDSavitch2001">(Savitch, 2001)</a> and <a xmlns=3D""=20
href=3D"#IDLamport1994">(Lamport, 1994)</a>, for it to work on.  That=20
should suffice.</p></div><div class=3D"section" lang=3D"en"=20
xml:lang=3D"en"><div class=3D"titlepage"><div><div><h2 class=3D"title"=20
style=3D"clear: both"><a id=3D"problem-statement"></a>Statement of the=20
problem</h2></div></div><div></div></div><p>Some weeks ago I decided to=20
plunge headlong into DocBook XML.  With no prior experience I immersed=20
myself in XSLT processors, FO processors, XML, the DocBook=20
specifications, catalogs (argh!) and, finally, refdb.  As I proceeded I=20
had a test document which steadily grew in complexity.  When I reached=20
the point of being able to ouptut pdf, I noticed something strange about=20
the Table of Contents.</p><p>What I noticed was that there was an entry=20
at the correct place.  It gave page number and leader dots.  What was=20
absent, however, were the words =E2Reference List=E2 at the start of the=20
entry.  I notified <span class=3D"emphasis"><em>refdb-users</em></span>=20
and continued.  My test document grew increasingly more complex as I=20
learned more.  Then I saw an amazing thing.  The entry had corrected=20
itself!  Yes, there was a pristine Table of Contents entry right where=20
it should have been all along.</p><p>My current working theory is that=20
it wil be absent in simple documents.  This doument is a test of that=20
theory.  So, here goes =E2=A6</p></div><div xmlns=3D"" id=3D"id2502984"=20
class=3D"bibliography"><div xmlns=3D"http://www.w3.org/1999/xhtml"=20
class=3D"titlepage"><div><div><h2 class=3D"title"><a=20
id=3D"id2502984"></a>Reference=20
List</h2></div></div><div></div></div><p><div id=3D"IDLamport1994"=20
class=3D"bibliomixed">Lamport L., 1994, LaTeX: A Document Preparation=20
System (Addison-Wesley, Massachusetts) </div></p><p><div=20
id=3D"IDSavitch2001" class=3D"bibliomixed">Savitch W., 2001, Java. An=20
Introduction to Computer Science and Programming (Prentice Hall, New=20
Jersey) </div></p><p><div id=3D"id2563143"=20
class=3D"bibliomixed"></div></p></div></div></body></html>
.........................................................................=
..............................................................

That's almost impossible to parse with the naked eye.  So, here is the=20
last part (the bibliography) in formatted form:

.........................................................................=
..............................................................
<... elision ...>
            <div xmlns=3D"" id=3D"id2502984" class=3D"bibliography">
                <div xmlns=3D"http://www.w3.org/1999/xhtml" class=3D"titl=
epage">
                    <div>
                        <div>
                            <h2 class=3D"title"><a=20
id=3D"id2502984"></a>Reference List</h2>
                        </div>
                    </div>
                    <div></div>
                </div>
                <p>
                    <div id=3D"IDLamport1994" class=3D"bibliomixed">
                        Lamport L., 1994, LaTeX: A Document Preparation=20
System (Addison-Wesley, Massachusetts)
                    </div>
                </p>
                <p>
                    <div id=3D"IDSavitch2001" class=3D"bibliomixed">
                        Savitch W., 2001, Java. An Introduction to=20
Computer Science and Programming (Prentice Hall, New Jersey)
                    </div>
                </p>
                <p>
                    <div id=3D"id2563143" class=3D"bibliomixed"></div>
                </p>
            </div>
        </div>
    </body>
</html>
.........................................................................=
..............................................................

It seems correct to me: each bibliography entry is enclosed in both=20
<p></p> and <div></div> elements.  Unless other eyes have spotted=20
something, I'm forced to agree this is a bug in the latest firebird and=20
firefox browsers on Debian.  Still, seems unusual that the remainder of=20
the document displays correctly.  Markus, perhaps you'd care to cut and=20
paste the xhtml into a file and see whether it renders correctly in your=20
browser?

As a final step I'll update to the latest CVS refdb and see if that has=20
any effect (shouldn't, since the bug appears to be in the browsers).

Regards,
David.

Re: [Refdb-users] xhtml references don't start on new line

From: David N. <dav...@bi...> - 2004-03-11 13:05:25

Hi Markus,

>I'm afraid this may be a display problem of your browser. The output
>looks just fine over here (Firebird 0.7 on FreeBSD).
>  
>

Have you tried xhtml output from saxon?  It's a bit gruesome.  Here is 
the result of processing my test document:
.........................................................................................................................................
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE html
  PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml"><head><title>TOC Reference 
List tester</title><meta name="generator" content="DocBook XSL 
Stylesheets V1.64.1"/></head><body><div class="article" lang="en"><div 
class="titlepage"><div><div><h2 class="title"><a id="d0e1"/>TOC 
Reference List tester</h2></div><div><div class="author"><h3 
class="author"><span class="firstname">David</span> <span 
class="surname">Nebauer</span></h3><div class="affiliation"><div 
class="address"><p><br/>
                    <tt xmlns:.0="" .0:class="email">&lt;<a 
.0:href="mailto:dav...@bi...">dav...@bi...</a>&gt;</tt><br/>
                </p></div></div></div></div><div><div 
class="revhistory"><table border="1" width="100%" summary="Revision 
history"><tr><th align="left" valign="top" colspan="3"><b>Revision 
History</b></th></tr><tr><td align="left">Revision 1</td><td 
align="left">2004-03-03</td><td align="left">DN</td></tr><tr><td 
align="left" colspan="3">Create 
document</td></tr></table></div></div></div><div/><hr/></div><div 
class="toc"><p><b>Table of Contents</b></p><dl><dt><span 
class="section"><a href="#spurious-quotes">Some 
quotes</a></span></dt><dt><span class="section"><a 
href="#problem-statement">Statement of the 
problem</a></span></dt><dt><span class="bibliography"><a 
href="#d0e54">Reference List</a></span></dt></dl></div><div 
class="section" lang="en"><div class="titlepage"><div><div><h2 
class="title" style="clear: both"><a id="spurious-quotes"/>Some 
quotes</h2></div></div><div/></div><p>In order to interest refdb in my 
document, I have to make some quotes, such as <a xmlns="" 
href="#IDSavitch2001">(Savitch, 2001)</a> and <a xmlns="" 
href="#IDLamport1994">(Lamport, 1994)</a>, for it to work on.  That 
should suffice.</p></div><div class="section" lang="en"><div 
class="titlepage"><div><div><h2 class="title" style="clear: both"><a 
id="problem-statement"/>Statement of the 
problem</h2></div></div><div/></div><p>Some weeks ago I decided to 
plunge headlong into DocBook XML.  With no prior experience I immersed 
myself in XSLT processors, FO processors, XML, the DocBook 
specifications, catalogs (argh!) and, finally, refdb.  As I proceeded I 
had a test document which steadily grew in complexity.  When I reached 
the point of being able to ouptut pdf, I noticed something strange about 
the Table of Contents.</p><p>What I noticed was that there was an entry 
at the correct place.  It gave page number and leader dots.  What was 
absent, however, were the words "Reference List" at the start of the 
entry.  I notified <span xmlns:.0="" 
.0:class="emphasis"><em>refdb-users</em></span> and continued.  My test 
document grew increasingly more complex as I learned more.  Then I saw 
an amazing thing.  The entry had corrected itself!  Yes, there was a 
pristine Table of Contents entry right where it should have been all 
along.</p><p>My current working theory is that it wil be absent in 
simple documents.  This doument is a test of that theory.  So, here goes 
...</p></div><div xmlns="" id="d0e54" class="bibliography"><div 
xmlns="http://www.w3.org/1999/xhtml" class="titlepage"><div><div><h2 
class="title"><a id="d0e54"/>Reference 
List</h2></div></div><div/></div><p><div id="IDLamport1994" 
class="bibliomixed">Lamport L., 1994, LaTeX: A Document Preparation 
System (Addison-Wesley, Massachusetts) </div></p><p><div 
id="IDSavitch2001" class="bibliomixed">Savitch W., 2001, Java. An 
Introduction to Computer Science and Programming (Prentice Hall, New 
Jersey) </div></p><p><div id="d0e185" 
class="bibliomixed"/></p></div></div></body></html>
.........................................................................................................................................

Note the ".0"s sprinkled throughout.  Here is the W3C validator's findings:
.........................................................................................................................................
This page is not Valid XHTML 1.0 Transitional!

Below are the results of attempting to parse this document with an SGML 
parser.

   1. Line 5, column 18: there is no attribute "xmlns

       <tt xmlns:.0="" .0:class="email">&lt;<a 
.0:href="mailto:davidnebauer@bigfoo
                    ^

   2. Line 5, column 21: ".0

       <tt xmlns:.0="" .0:class="email">&lt;<a 
.0:href="mailto:davidnebauer@bigfoo
                       ^

   3. Line 5, column 29: an attribute specification must start with a 
name or name token (explain...).

       <tt xmlns:.0="" .0:class="email">&lt;<a 
.0:href="mailto:davidnebauer@bigfoo
                               ^

   4. Line 5, column 45: ".0

  ... xmlns:.0="" .0:class="email">&lt;<a 
.0:href="mailto:dav...@bi...
                                          ^

   5. Line 5, column 52: an attribute specification must start with a 
name or name token (explain...).

  ....0="" .0:class="email">&lt;<a 
.0:href="mailto:dav...@bi...">david
                                          ^

   6. Line 6, column 970: there is no attribute "xmlns" (explain...).

  ...o make some quotes, such as <a xmlns="" 
href="#IDSavitch2001">(Savitch, 2001)
                                          ^

   7. Line 6, column 1928: there is no attribute "xmlns

  ...of the entry.  I notified <span xmlns:.0="" 
.0:class="emphasis"><em>refdb-use
                                          ^

   8. Line 6, column 1931: ".0

  ...the entry.  I notified <span xmlns:.0="" 
.0:class="emphasis"><em>refdb-users<
                                          ^

   9. Line 6, column 1939: an attribute specification must start with a 
name or name token (explain...).

  ...y.  I notified <span xmlns:.0="" 
.0:class="emphasis"><em>refdb-users</em></sp
                                          ^

  10. Line 6, column 2374: there is no attribute "xmlns" (explain...).

  ...ry.  So, here goes ?</p></div><div xmlns="" id="d0e54" 
class="bibliography"
                                          ^

  11. Line 6, column 2504: ID "d0e54" already defined (explain...).

  ...lepage"><div><div><h2 class="title"><a id="d0e54"/>Reference 
List</h2></div><
                                          ^

  12. Line 6, column 2381: ID "d0e54" first defined here (explain...).

  ..., here goes ?</p></div><div xmlns="" id="d0e54" 
class="bibliography"><div x
                                          ^

  13. Line 6, column 2601: document type does not allow element "div" 
here; missing one of "object", "applet", "map", "iframe", "button", 
"ins", "del" start-tag

  ...<div id="IDLamport1994" class="bibliomixed">Lamport L., 1994, 
LaTeX: A Docume
                                          ^

  14. Line 6, column 2745: document type does not allow element "div" 
here; missing one of "object", "applet", "map", "iframe", "button", 
"ins", "del" start-tag

  ...<div id="IDSavitch2001" class="bibliomixed">Savitch W., 2001, Java. 
An Introd
                                          ^

  15. Line 6, column 2900: document type does not allow element "div" 
here; missing one of "object", "applet", "map", "iframe", "button", 
"ins", "del" start-tag

  .../p><p><div id="d0e185" 
class="bibliomixed"/></p></div></div></body></html>
                                          ^
.........................................................................................................................................

Note: Saxon continues to produce pristine HTML and PDF.

Xalan gives similar results to xsltproc.  I've no idea what it means, 
but Saxon's output may give a clue to what is happening with xhtml 
translation generally.

Regards,
David.

Re: [Refdb-users] xhtml references don't start on new line

From: Markus H. <mar...@mh...> - 2004-03-11 18:35:50

David Nebauer writes:
 > Have you tried xhtml output from saxon?  It's a bit gruesome.  Here is 
 > the result of processing my test document:
 > .........................................................................................................................................

Yes I did. There were no such problems with ".0' being randomly
inserted. However, saxon's output shows the empty xmlns attribute, in
contrast to my version of xsltproc:

markus@~/sgml/dtd# xsltproc --version
Using libxml 20428, libxslt 10023 and libexslt 714
xsltproc was compiled against libxml 20428, libxslt 10023 and libexslt 714
libxslt 10023 was compiled against libxml 20428
libexslt 714 was compiled against libxml 20428

I have no clue yet what this means, but with saxon showing the xmlns
problem I do have a chance to investigate this.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [Refdb-users] xhtml references don't start on new line

From: Markus H. <mar...@mh...> - 2004-03-11 19:34:04

Markus Hoenicka writes:
 > I have no clue yet what this means, but with saxon showing the xmlns
 > problem I do have a chance to investigate this.
 > 

Googling finally gave me a clue. My driver file did not define a
default namespace, therefore some processors felt the need to insert
an empty namespace (which is legal xhtml but it obviously confuses
browsers). This is fixed now. I've also removed the <p> elements
wrapping each reference as the <div> elements are sufficient. I've
finally updated the bibliography template to match the latest DocBook
version which fixed the duplicate id bug. Please have a look at
version 1.5 of docbk-refdb-xhtml.xsl.in. This should ease your pain.

validator.w3.org now tells me:
This Page Is Valid XHTML 1.0 Transitional!

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [Refdb-users] xhtml references don't start on new line

From: David N. <dav...@bi...> - 2004-03-12 10:56:57

Markus Hoenicka wrote:

>markus@~/sgml/dtd# xsltproc --version
>Using libxml 20428, libxslt 10023 and libexslt 714
>xsltproc was compiled against libxml 20428, libxslt 10023 and libexslt 714
>libxslt 10023 was compiled against libxml 20428
>libexslt 714 was compiled against libxml 20428
>
Interesting that xsltproc on my system produced the 'xmlns' attribute 
while it did not on your version.

david@abelard:ref-toc$ xsltproc --version
Using libxml 20606, libxslt 10102 and libexslt 802
xsltproc was compiled against libxml 20603, libxslt 10102 and libexslt 802
libxslt 10102 was compiled against libxml 20603
libexslt 802 was compiled against libxml 20603

Of course, it may be that the packagers for our respective platforms 
compiled xsltproc with different options.

I see also that Saxon gave different output on each system.

Here are the versions of the respective processors:
xalan2 (docbook extensions version 1.64.1.0-4.2)
saxon (docbook extensions version 1.64.1.0-4.2, appears to be 6.4.4)
xt (debian package libxt-java version 0.20020426a-2 -- uncertain of the 
actual version of xt itself)

Here are my results with the latest cvs refdb, testing the output with 
w3 validator and for display in firefox:

processor   firefox display   w3 validator
---------   ---------------   ------------
xsltproc          OK               OK
xt                OK               OK
xalan             OK            Error (1)
saxon           Error (2)       Error (3)
            

Error (1): xalan -- w3 validator

The w3 validator found the following problems:
.......................................................................................................................
This page is not Valid XHTML 1.0 Transitional!

Below are the results of attempting to parse this document with an SGML 
parser.

   1. Line 3, column 481: there is no attribute "xmlns" (explain...).

  ...ation"><div class="address"><p 
xmlns="http://www.w3.org/1999/xhtml"><span sty
                                          ^

   2. Line 5, column 428: there is no attribute "xmlns" (explain...).

  .../><hr /></div><div class="toc" 
xmlns="http://www.w3.org/1999/xhtml"><p><b>Tab
                                          ^
.......................................................................................................................

xalan produced the following output file:
.......................................................................................................................
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml"><head><title>TOC Reference 
List tester</title><meta content="DocBook XSL Stylesheets V1.64.1" 
name="generator" /></head><body><div class="article" lang="en"><div 
class="titlepage"><div><div><h2 class="title"><a id="N10001" />TOC 
Reference List tester</h2></div><div><div class="author"><h3 
class="author"><span class="firstname">David</span> <span 
class="surname">Nebauer</span></h3><div class="affiliation"><div 
class="address"><p xmlns="http://www.w3.org/1999/xhtml"><span 
style="white-space: pre;">
                    <tt class="email">&lt;<a 
href="mailto:dav...@bi...">dav...@bi...</a>&gt;</tt>
                </span></p></div></div></div></div><div><div 
class="revhistory"><table summary="Revision history" width="100%" 
border="1"><tr><th colspan="3" valign="top" align="left"><b>Revision 
History</b></th></tr><tr><td align="left">Revision 1</td><td 
align="left">2004-03-03</td><td align="left">DN</td></tr><tr><td 
colspan="3" align="left">Create 
document</td></tr></table></div></div></div><div /><hr /></div><div 
class="toc" xmlns="http://www.w3.org/1999/xhtml"><p><b>Table of 
Contents</b></p><dl><dt><span class="section"><a 
href="#spurious-quotes">Some quotes</a></span></dt><dt><span 
class="section"><a href="#problem-statement">Statement of the 
problem</a></span></dt><dt><span class="bibliography"><a 
href="#N1003E">Reference List</a></span></dt></dl></div><div 
class="section" lang="en"><div class="titlepage"><div><div><h2 
class="title" style="clear: both"><a id="spurious-quotes" />Some 
quotes</h2></div></div><div /></div><p 
xmlns="http://www.w3.org/1999/xhtml">In order to interest refdb in my 
document, I have to make some quotes, such as <a 
href="#IDSavitch2001">(Savitch, 2001)</a> and <a 
href="#IDLamport1994">(Lamport, 1994)</a>, for it to work on.  That 
should suffice.</p></div><div class="section" lang="en"><div 
class="titlepage"><div><div><h2 class="title" style="clear: both"><a 
id="problem-statement" />Statement of the problem</h2></div></div><div 
/></div><p xmlns="http://www.w3.org/1999/xhtml">Some weeks ago I decided 
to plunge headlong into DocBook XML.  With no prior experience I 
immersed myself in XSLT processors, FO processors, XML, the DocBook 
specifications, catalogs (argh!) and, finally, refdb.  As I proceeded I 
had a test document which steadily grew in complexity.  When I reached 
the point of being able to ouptut pdf, I noticed something strange about 
the Table of Contents.</p><p xmlns="http://www.w3.org/1999/xhtml">What I 
noticed was that there was an entry at the correct place.  It gave page 
number and leader dots.  What was absent, however, were the words 
"Reference List" at the start of the entry.  I notified <span 
class="emphasis"><em>refdb-users</em></span> and continued.  My test 
document grew increasingly more complex as I learned more.  Then I saw 
an amazing thing.  The entry had corrected itself!  Yes, there was a 
pristine Table of Contents entry right where it should have been all 
along.</p><p xmlns="http://www.w3.org/1999/xhtml">My current working 
theory is that it wil be absent in simple documents.  This doument is a 
test of that theory.  So, here goes ...</p></div><div 
class="bibliography"><div class="titlepage"><div><div><h2 
class="title"><a id="N1003E" />Reference List</h2></div></div><div 
/></div><div class="bibliomixed" id="IDLamport1994">Lamport L., 1994, 
LaTeX: A Document Preparation System (Addison-Wesley, Massachusetts) 
</div><div class="bibliomixed" id="IDSavitch2001">Savitch W., 2001, 
Java. An Introduction to Computer Science and Programming (Prentice 
Hall, New Jersey) </div><div class="bibliomixed" id="N100FE" 
/></div></div></body></html>
.......................................................................................................................


Error (2): Saxon -- Firefox display error

Firefox unable to display.  Displayed the following:
.......................................................................................................................
XML Parsing Error: not well-formed
Location: 
file:///home/david/data/computing/docbook-xml/learn/ref-toc/reftoc.xhtml
Line Number 5, Column 22:
<tt xmlns:.0="" .0:class="email">&lt;<a 
.0:href="mailto:dav...@bi...">dav...@bi...</a>&gt;</tt><br/> 
---------------------^
.......................................................................................................................



Error (3): Saxon -- w3 validator

w3 validator complained of:
.......................................................................................................................
This page is not Valid XHTML 1.0 Transitional!

Below are the results of attempting to parse this document with an SGML 
parser.

   1. Line 5, column 18: there is no attribute "xmlns

       <tt xmlns:.0="" .0:class="email">&lt;<a 
.0:href="mailto:davidnebauer@bigfoo
                    ^

   2. Line 5, column 21: ".0

       <tt xmlns:.0="" .0:class="email">&lt;<a 
.0:href="mailto:davidnebauer@bigfoo
                       ^

   3. Line 5, column 29: an attribute specification must start with a 
name or name token (explain...).

       <tt xmlns:.0="" .0:class="email">&lt;<a 
.0:href="mailto:davidnebauer@bigfoo
                               ^

   4. Line 5, column 45: ".0

  ... xmlns:.0="" .0:class="email">&lt;<a 
.0:href="mailto:dav...@bi...
                                          ^

   5. Line 5, column 52: an attribute specification must start with a 
name or name token (explain...).

  ....0="" .0:class="email">&lt;<a 
.0:href="mailto:dav...@bi...">david
                                          ^

   6. Line 6, column 973: there is no attribute "xmlns

  ...ake some quotes, such as <a xmlns:.0="" 
.0:href="#IDSavitch2001">(Savitch, 20
                                          ^

   7. Line 6, column 976: ".0

  ... some quotes, such as <a xmlns:.0="" 
.0:href="#IDSavitch2001">(Savitch, 2001)
                                          ^

   8. Line 6, column 983: an attribute specification must start with a 
name or name token (explain...).

  ...uotes, such as <a xmlns:.0="" .0:href="#IDSavitch2001">(Savitch, 
2001)</a> an
                                          ^

   9. Line 6, column 1040: ".0

  ...avitch, 2001)</a> and <a xmlns:.0="" 
.0:href="#IDLamport1994">(Lamport, 1994)
                                          ^

  10. Line 6, column 1047: an attribute specification must start with a 
name or name token (explain...).

  ... 2001)</a> and <a xmlns:.0="" .0:href="#IDLamport1994">(Lamport, 
1994)</a>, f
                                          ^

  11. Line 6, column 1940: there is no attribute "xmlns

  ...of the entry.  I notified <span xmlns:.0="" 
.0:class="emphasis"><em>refdb-use
                                          ^

  12. Line 6, column 1943: ".0

  ...the entry.  I notified <span xmlns:.0="" 
.0:class="emphasis"><em>refdb-users<
                                          ^

  13. Line 6, column 1951: an attribute specification must start with a 
name or name token (explain...).

  ...y.  I notified <span xmlns:.0="" 
.0:class="emphasis"><em>refdb-users</em></sp
                                          ^
.......................................................................................................................

The xhtml file is:
.......................................................................................................................
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE html
  PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml"><head><title>TOC Reference 
List tester</title><meta name="generator" content="DocBook XSL 
Stylesheets V1.64.1"/></head><body><div class="article" lang="en"><div 
class="titlepage"><div><div><h2 class="title"><a id="d0e1"/>TOC 
Reference List tester</h2></div><div><div class="author"><h3 
class="author"><span class="firstname">David</span> <span 
class="surname">Nebauer</span></h3><div class="affiliation"><div 
class="address"><p><br/>
                    <tt xmlns:.0="" .0:class="email">&lt;<a 
.0:href="mailto:dav...@bi...">dav...@bi...</a>&gt;</tt><br/>
                </p></div></div></div></div><div><div 
class="revhistory"><table border="1" width="100%" summary="Revision 
history"><tr><th align="left" valign="top" colspan="3"><b>Revision 
History</b></th></tr><tr><td align="left">Revision 1</td><td 
align="left">2004-03-03</td><td align="left">DN</td></tr><tr><td 
align="left" colspan="3">Create 
document</td></tr></table></div></div></div><div/><hr/></div><div 
class="toc"><p><b>Table of Contents</b></p><dl><dt><span 
class="section"><a href="#spurious-quotes">Some 
quotes</a></span></dt><dt><span class="section"><a 
href="#problem-statement">Statement of the 
problem</a></span></dt><dt><span class="bibliography"><a 
href="#d0e54">Reference List</a></span></dt></dl></div><div 
class="section" lang="en"><div class="titlepage"><div><div><h2 
class="title" style="clear: both"><a id="spurious-quotes"/>Some 
quotes</h2></div></div><div/></div><p>In order to interest refdb in my 
document, I have to make some quotes, such as <a xmlns:.0="" 
.0:href="#IDSavitch2001">(Savitch, 2001)</a> and <a xmlns:.0="" 
.0:href="#IDLamport1994">(Lamport, 1994)</a>, for it to work on.  That 
should suffice.</p></div><div class="section" lang="en"><div 
class="titlepage"><div><div><h2 class="title" style="clear: both"><a 
id="problem-statement"/>Statement of the 
problem</h2></div></div><div/></div><p>Some weeks ago I decided to 
plunge headlong into DocBook XML.  With no prior experience I immersed 
myself in XSLT processors, FO processors, XML, the DocBook 
specifications, catalogs (argh!) and, finally, refdb.  As I proceeded I 
had a test document which steadily grew in complexity.  When I reached 
the point of being able to ouptut pdf, I noticed something strange about 
the Table of Contents.</p><p>What I noticed was that there was an entry 
at the correct place.  It gave page number and leader dots.  What was 
absent, however, were the words "Reference List" at the start of the 
entry.  I notified <span xmlns:.0="" 
.0:class="emphasis"><em>refdb-users</em></span> and continued.  My test 
document grew increasingly more complex as I learned more.  Then I saw 
an amazing thing.  The entry had corrected itself!  Yes, there was a 
pristine Table of Contents entry right where it should have been all 
along.</p><p>My current working theory is that it wil be absent in 
simple documents.  This doument is a test of that theory.  So, here goes 
...</p></div><div class="bibliography"><div 
class="titlepage"><div><div><h2 class="title"><a id="d0e54"/>Reference 
List</h2></div></div><div/></div><div id="IDLamport1994" 
class="bibliomixed">Lamport L., 1994, LaTeX: A Document Preparation 
System (Addison-Wesley, Massachusetts) </div><div id="IDSavitch2001" 
class="bibliomixed">Savitch W., 2001, Java. An Introduction to Computer 
Science and Programming (Prentice Hall, New Jersey) </div><div 
id="d0e185" class="bibliomixed"/></div></div></body></html>
.......................................................................................................................

The versions of Saxon and Xalan used initially were the docbook 
extensions versions.  I then modified the classpaths to force the use of 
the "standard" versions.  The exact same outcomes occurred.

For reference, here is the (error-free) output from xsltproc (for anyone 
still reading, apologies for the length of this email):
.......................................................................................................................
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" 
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml"><head><meta 
http-equiv="Content-Type" content="text/html; charset=UTF-8" 
/><title>TOC Reference List tester</title><meta name="generator" 
content="DocBook XSL Stylesheets V1.64.1" /></head><body><div 
class="article" lang="en" xml:lang="en"><div 
class="titlepage"><div><div><h2 class="title"><a id="id2405464"></a>TOC 
Reference List tester</h2></div><div><div class="author"><h3 
class="author"><span class="firstname">David</span> <span 
class="surname">Nebauer</span></h3><div class="affiliation"><div 
class="address"><p><br />
                    <tt class="email">&lt;<a 
href="mailto:dav...@bi...">dav...@bi...</a>&gt;</tt><br 
/>
                </p></div></div></div></div><div><div 
class="revhistory"><table border="1" width="100%" summary="Revision 
history"><tr><th align="left" valign="top" colspan="3"><b>Revision 
History</b></th></tr><tr><td align="left">Revision 1</td><td 
align="left">2004-03-03</td><td align="left">DN</td></tr><tr><td 
align="left" colspan="3">Create 
document</td></tr></table></div></div></div><div></div><hr /></div><div 
class="toc"><p><b>Table of Contents</b></p><dl><dt><span 
class="section"><a href="#spurious-quotes">Some 
quotes</a></span></dt><dt><span class="section"><a 
href="#problem-statement">Statement of the 
problem</a></span></dt><dt><span class="bibliography"><a 
href="#id2505208">Reference List</a></span></dt></dl></div><div 
class="section" lang="en" xml:lang="en"><div 
class="titlepage"><div><div><h2 class="title" style="clear: both"><a 
id="spurious-quotes"></a>Some 
quotes</h2></div></div><div></div></div><p>In order to interest refdb in 
my document, I have to make some quotes, such as <a 
href="#IDSavitch2001">(Savitch, 2001)</a> and <a 
href="#IDLamport1994">(Lamport, 1994)</a>, for it to work on.  That 
should suffice.</p></div><div class="section" lang="en" 
xml:lang="en"><div class="titlepage"><div><div><h2 class="title" 
style="clear: both"><a id="problem-statement"></a>Statement of the 
problem</h2></div></div><div></div></div><p>Some weeks ago I decided to 
plunge headlong into DocBook XML.  With no prior experience I immersed 
myself in XSLT processors, FO processors, XML, the DocBook 
specifications, catalogs (argh!) and, finally, refdb.  As I proceeded I 
had a test document which steadily grew in complexity.  When I reached 
the point of being able to ouptut pdf, I noticed something strange about 
the Table of Contents.</p><p>What I noticed was that there was an entry 
at the correct place.  It gave page number and leader dots.  What was 
absent, however, were the words "Reference List" at the start of the 
entry.  I notified <span class="emphasis"><em>refdb-users</em></span> 
and continued.  My test document grew increasingly more complex as I 
learned more.  Then I saw an amazing thing.  The entry had corrected 
itself!  Yes, there was a pristine Table of Contents entry right where 
it should have been all along.</p><p>My current working theory is that 
it wil be absent in simple documents.  This doument is a test of that 
theory.  So, here goes ...</p></div><div class="bibliography"><div 
class="titlepage"><div><div><h2 class="title"><a 
id="id2505208"></a>Reference List</h2></div></div><div></div></div><div 
id="IDLamport1994" class="bibliomixed">Lamport L., 1994, LaTeX: A 
Document Preparation System (Addison-Wesley, Massachusetts) </div><div 
id="IDSavitch2001" class="bibliomixed">Savitch W., 2001, Java. An 
Introduction to Computer Science and Programming (Prentice Hall, New 
Jersey) </div><div id="id2565363" 
class="bibliomixed"></div></div></div></body></html>
.......................................................................................................................


In summary, xalan is putting a xmlns attribute into:
- the address <p> element, and
- the toc <div> element.

The XHTML standard does not list xmlns as an attribute of either of 
these elements.  Unless it is a common attribute that can occur 
anywhere, this is invalid XHTML despite displaying correctly in Firefox.

In summary, saxon is putting xmlns attributes and ".0"s in <tt>, <a> and 
<span> elements.

So, what happens if I process the source document using the standard 
docbook stylesheets (without the refdb stylesheets)?  Aha!  Both xalan 
and saxon returned the same errors as they did with the refdb stylesheets.

Ergo, these remaining problems relate to saxon and xalan in translating 
xhtml, _not_ to refdb.  More precisely, when using the standard docbook 
stylesheets.  More precisely yet, they relate to the versions of saxon 
and xalan installed on Debian Sarge, since I gather Markus did not get 
the exact same errors as I did.

Now for a spot of googling.

First: Saxon.

At 
<http://lists.oasis-open.org/archives/docbook-apps/200203/msg00216.html> 
we find no less a luminary than Norm Walsh explaining that the Saxon 
problem is due to a bug which will be fixed in the next Saxon release.  
Obviously, Markus' version of Saxon includes the bug fix.  Amazing that 
the Debian Sarge version is so out of date ...

Next: Xalan.

I found a newsgroup thread that _appears_ to address this issue.  The 
url for this site is appallingly long: 
<http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&threadm=blcbnk%24h4f%2402%241%40news.t-online.com&rnum=1&prev=/groups%3Fq%3Dxalan%2B%2522%253Cp%2Bxmlns%253Dhttp://www.w3.org/1999/xhtml%253E%2522%26hl%3Den%26lr%3D%26ie%3DUTF-8%26selm%3Dblcbnk%2524h4f%252402%25241%2540news.t-online.com%26rnum%3D1>.
The solutions discussed are a bit beyond my knowledge level at the 
moment, but may make sense to others.  The main answer points the 
questionner to <http://www.topxml.com/people/bosley/defaultns.asp>.

So, thanks Markus for fixing the problems so far.  Looks like I need an 
up-to-date Saxon.

Regards,
David.

Re: [Refdb-users] xhtml references don't start on new line

From: David N. <dav...@bi...> - 2004-03-12 13:39:02

David Nebauer wrote:

> At 
> <http://lists.oasis-open.org/archives/docbook-apps/200203/msg00216.html> 
> we find no less a luminary than Norm Walsh explaining that the Saxon 
> problem is due to a bug which will be fixed in the next Saxon 
> release.  Obviously, Markus' version of Saxon includes the bug fix.  
> Amazing that the Debian Sarge version is so out of date ...
>
> So, thanks Markus for fixing the problems so far.  Looks like I need 
> an up-to-date Saxon.

Yep, Saxon 6.5.4 renders xhtml perfectly where 6.4.4 did not.

For those of you wondering why I wasn't running 6.5 in the first place, 
and with 30 seconds to spare, prepare to be amazed...

(This assumes a passing familiarity with Debian.)

The package lib-saxon-java contains the saxon processor.  It was frozen 
at 6.4.4 for reasons known only to it's packager.  The bug reports are 
full of pleas to upgrade it to 6.5.x so that, wait for it, people could 
render xhtml.  Unfortunately, the maintainer simply refused.

In response, a new package called libsaxon-java was created (in case you 
missed it, they removed the first hyphen).  It tracks the latest stable 
version.  This package was put in 'contrib', not 'main'.  And guess who 
didn't have contrib listed in his sources.list file and therefore never 
saw it?

While I love it dearly, this kind of thing could only happen on Planet 
Debian.

For those of you who followed me on this, I hope you enjoyed the ride.  
At least we found a few bugs along the way.

Regards,
David.

Re: [Refdb-users] xhtml references don't start on new line

From: David N. <dav...@bi...> - 2004-03-11 12:57:35

Hi Markus,

>I'm afraid this may be a display problem of your browser. The output
>looks just fine over here (Firebird 0.7 on FreeBSD).
>

Updating to latest CVS version had no effect, as I suspected.

I have, however, noted another rendering bug: the actual references do 
not appear as hyperlinks in Firebird or Firefox.

The offending xhtml fragment is:
........................................................................................................................
... quotes, such as <a xmlns="" href="#IDSavitch2001">(Savitch, 
2001)</a> and <a xmlns="" href="#IDLamport1994">(Lamport, 1994)</a>, for 
it to work ...
........................................................................................................................

The problem is the empty xmlns attributes.  If I manually remove them, 
both browsers display the hyperlinks.  If the namespace attribute is a 
legal part of the xhtml specification, then this is another problem with 
Firebird/Firefox and xhtml.

Almost as an afterthought, at this point I ran the original xhtml file 
through the W3C validator <http://validator.w3.org/>.  It returned the 
following information:
........................................................................................................................
This page is not Valid XHTML 1.0 Transitional!

Below are the results of attempting to parse this document with an SGML 
parser.

   1. Line 5, column 1002: there is no attribute "xmlns" (explain...).

  ...o make some quotes, such as <a xmlns="" 
href="#IDSavitch2001">(Savitch, 2001)
                                          ^

   2. Line 5, column 2413: there is no attribute "xmlns" (explain...).

  ...ry.  So, here goes ?</p></div><div xmlns="" id="id2502984" 
class="bibliogra
                                          ^

   3. Line 5, column 2547: ID "id2502984" already defined (explain...).

  ...lepage"><div><div><h2 class="title"><a id="id2502984"></a>Reference 
List</h2>
                                          ^

   4. Line 5, column 2420: ID "id2502984" first defined here (explain...).

  ..., here goes ?</p></div><div xmlns="" id="id2502984" 
class="bibliography"><d
                                          ^

   5. Line 5, column 2656: document type does not allow element "div" 
here; missing one of "object", "applet", "map", "iframe", "button", 
"ins", "del" start-tag

  ...<div id="IDLamport1994" class="bibliomixed">Lamport L., 1994, 
LaTeX: A Docume
                                          ^

   6. Line 5, column 2800: document type does not allow element "div" 
here; missing one of "object", "applet", "map", "iframe", "button", 
"ins", "del" start-tag

  ...<div id="IDSavitch2001" class="bibliomixed">Savitch W., 2001, Java. 
An Introd
                                          ^

   7. Line 5, column 2957: document type does not allow element "div" 
here; missing one of "object", "applet", "map", "iframe", "button", 
"ins", "del" start-tag

  ...><p><div id="id2563143" 
class="bibliomixed"></div></p></div></div></body></ht
                                          ^
........................................................................................................................

I note that this result is from an SGML parser and W3C list some 
limitations of that parser, but none of them seem to apply here.  The 
W3C validator is complaining about the parts of the document that render 
poorly on my Firefox and Firebird:
 - Errors 1 and 2 ("no attribute xmlns")
     -> the citations involved don't render as hyperlinks.
 - Errors 5-7 ("document type does not allow element 'div' here")
     -> these are the references, which don't start on their own line.

I had a look at the XHTML standard as parsed by dtdparse (see 
<http://dtdparse.sourceforge.net/xhtml10/>) and found that:
- the <a> element has no 'xmlns' attribute.  Perhaps this is a wandering 
namespace declaration?
- the <p> element cannot have a <div> element as a child (nor can a 
<div> element have a <p> element as a parent).

Is it possible my browsers have it right after all?

Regards,
David

Re: [Refdb-users] xhtml references don't start on new line

From: Markus H. <mar...@mh...> - 2004-03-11 18:35:39

David Nebauer writes:
 > I have, however, noted another rendering bug: the actual references do 
 > not appear as hyperlinks in Firebird or Firefox.
 > 
 > The offending xhtml fragment is:
 > ........................................................................................................................
 > ... quotes, such as <a xmlns="" href="#IDSavitch2001">(Savitch, 
 > 2001)</a> and <a xmlns="" href="#IDLamport1994">(Lamport, 1994)</a>, for 
 > it to work ...
 > ........................................................................................................................
 > 
 > The problem is the empty xmlns attributes.  If I manually remove them, 
 > both browsers display the hyperlinks.  If the namespace attribute is a 
 > legal part of the xhtml specification, then this is another problem with 
 > Firebird/Firefox and xhtml.
 > 

Now this really gets me worried. First of all, the citations in a test
document processed on my system (using xsltproc) look like this:

<p>We cite this one <a href="#IDMiller1999">
(1)</a> and <a href="#IDWalsh1999">(2)</a> and <a
href="#IDPhadke1994">(3)</a></p>

That is, there is no xmlns attribute.

Second, if you look at the xhtml stylesheets, there is no code
whatsoever to insert such an attribute. The relevant instructions that
create the links look like this:

      <a href='#{$realtarget}'>

Third, if you look at the xhtml dtd, there is no such attribute at
all.

I somehow suspect that the xslt processor that you used thought it to
be a good idea to have an xmlns attribute, but I'm at a loss how this
could be suppressed.

 > Almost as an afterthought, at this point I ran the original xhtml file 
 > through the W3C validator <http://validator.w3.org/>.  It returned the 
 > following information:

My bad. I should have done this in the first place.

 >    5. Line 5, column 2656: document type does not allow element "div" 
 > here; missing one of "object", "applet", "map", "iframe", "button", 
 > "ins", "del" start-tag
 > 
 >   ...<div id="IDLamport1994" class="bibliomixed">Lamport L., 1994, 
 > LaTeX: A Docume
 >                                           ^

This indicates that the refdb xhtml driver file creates invalid xhtml
output which needs to be fixed.

 > Is it possible my browsers have it right after all?
 > 

Absolutely. I apologize for jumping the gun. I'll do my part to fix
the problem, but the xmlns thing might be out of my reach.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [Refdb-users] xhtml references don't start on new line

From: David N. <dav...@bi...> - 2004-03-11 13:08:43

Markus Hoenicka wrote:

>I'm afraid this may be a display problem of your browser. The output
>looks just fine over here (Firebird 0.7 on FreeBSD).
>  
>

In doing some troubleshooting with Saxon, it complained about 
'ADVSTURLSTYLE' in docbk-refdb-xhtml.xsl, line 75.

The 'T' needs to be removed as it was with the html and fo stylesheets.

Regards,
David.

Re: [Refdb-users] xhtml references don't start on new line

From: Markus H. <mar...@mh...> - 2004-03-11 18:35:48

David Nebauer writes:
 > Markus Hoenicka wrote:
 > 
 > >I'm afraid this may be a display problem of your browser. The output
 > >looks just fine over here (Firebird 0.7 on FreeBSD).
 > >  
 > >
 > 
 > In doing some troubleshooting with Saxon, it complained about 
 > 'ADVSTURLSTYLE' in docbk-refdb-xhtml.xsl, line 75.
 > 
 > The 'T' needs to be removed as it was with the html and fo stylesheets.
 > 

Thanks for reminding me. I've removed the 'T' from the html stylesheet
just yesterday but forgot to check the xhtml stylesheet as well.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [Refdb-users] xhtml references don't start on new line

From: Bruce D'A. <bd...@fa...> - 2004-03-11 14:10:49

On Mar 11, 2004, at 7:39 AM, David Nebauer wrote:

> I have, however, noted another rendering bug: the actual references do 
> not appear as hyperlinks in Firebird or Firefox.

Perhaps this highlights why it would be valuable to finish the php 
interface started by Rob and Matt?  In that case, rendering is all 
handled via XSLT and CSS, which would likely leave less room for 
errors.

Bruce

Re: [Refdb-users] xhtml references don't start on new line

From: Markus H. <mar...@mh...> - 2004-03-11 18:35:38

Bruce D'Arcus writes:
 > 
 > On Mar 11, 2004, at 7:39 AM, David Nebauer wrote:
 > 
 > > I have, however, noted another rendering bug: the actual references do 
 > > not appear as hyperlinks in Firebird or Firefox.
 > 
 > Perhaps this highlights why it would be valuable to finish the php 
 > interface started by Rob and Matt?  In that case, rendering is all 
 > handled via XSLT and CSS, which would likely leave less room for 
 > errors.
 > 

I don't see why. We're dealing with logical errors (e.g. the incorrect
nesting of <p> and <div>) and processor errors. We can't do anything
about the latter except using a different one. The former are
independent of the programming language used. You can create the same
incorrect output with any language. Errors in XSLT are only harder to
spot as that language has a very lousy signal-to-noise ratio.

It rather highlights the need for more thorough testing than I alone
can do. Thererore, thanks to all who run tests and report the
problems.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [Refdb-users] xhtml references don't start on new line

From: Markus H. <mar...@mh...> - 2004-03-11 19:39:47

Markus Hoenicka writes:
 > Bruce D'Arcus writes:
 >  > Perhaps this highlights why it would be valuable to finish the php 
 >  > interface started by Rob and Matt?  In that case, rendering is all 
 >  > handled via XSLT and CSS, which would likely leave less room for 
 >  > errors.
 >  > 
 > 
 > I don't see why. We're dealing with logical errors (e.g. the incorrect
 > nesting of <p> and <div>) and processor errors. We can't do anything
 > about the latter except using a different one. The former are
 > independent of the programming language used. You can create the same
 > incorrect output with any language. Errors in XSLT are only harder to
 > spot as that language has a very lousy signal-to-noise ratio.
 > 

Now that the problem is fixed, things look even worse: These were bugs in
the XSLT code that didn't have anything to do with the refdbd output.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de