Lars Vogel wrote on 2014-02-18:
By the way, even with this error calibre has not issues opening the
generate epub file. Maybe this is just a false alarm by epubcheck?
I doubt that, because it is a requirement of the EPUB spec, see:
Chris Ridd wrote:
...
It sounds like something has gone wrong with the zip file creation.
The first filename in the zip file is required to be called mimetype and it
must contain the characters application/epub+zip. I don't think that's
changed between epub 2 to epub 3.
1. Delete
Thomas Schraitle wrote, on 03.04.2013 10:41:
Currently, it's only available as HTML. However, if there are enough
readers who would find an EPUB version useful, I'll publish one. :)
+1 for the EPUB version!
Michael
-
To
Kerry, Richard wrote:
...
How do you implement stage 2. Copy any image files into the output
directory. ?
...
Is there a way in Ant of doing a copy based on data which only becomes
available during the build process ?
Try the following approach (untested):
- generate a file which contains
Lars Vogel wrote:
not sure if this is the right distribution list.
Probably no ;-)
Is their a way to tell Apache Ant to fail the build if such an exception
occurs?
No idea if this is will work but try a combination of ant-contribs
trycatch and (ant builtin) fail/ task (in the catch block).
Given a legacy DocBook 4.3 file, quite large (approx. 640 KB, composed
out of several XML files using XInclude).
This source file validates w/o any errors/warnings using xmllint.
Trying to generate XHTML with the DocBook XSL stylesheets V1.77.1
(*no* customization used, using xsltproc) shows
On 29.11.2012 14:21, Stefan Hinz wrote:
Since no one has replied so far, let me make a wild guess. Could it be
you're using xrefs that point to elements that have no title? For
example, something like this:
No, I don't use xrefs in the document.
In the meantime I'm almost convinced that it
Bob Stayton wrote, on 29.11.2012 19:10:
Indeed, it is a bug in the stylesheet. It was this combination that
showed the problem:
bridgeheadliteral\stepcounter{counter}/literal/bridgehead
The bridgehead is handled in mode=title.markup, but its template
looked like this:
xsl:template
Lars Vogel wrote:
During the conversion to .mobi via the Amazon command line tool
kindlegen I get several warnings which may indicate that the generated
epub has some issues. See below for the error messages. I attached the
...
I'd suggest to check the generated EPUB using EpubCheck
Thomas Schraitle wrote:
...
4. Can I see the source code?
Sure! The source code of this book is published and maintained as a
Mercurial repository. See the project page at
http://sf.net/projects/doccookbook/
Even with a regular SF account I don't have permissions to browse the
source code
Thomas Schraitle wrote:
5. Ok, convinced! Howe can i support the project?
...
* Send me your topics directly or use the project’s ticket system
(see https://sf.net/p/doccookbook/tickets/15/). If you want to use
the latter, you’ll have to register at Sourceforge
Write access required to
Hi Eric,
Eric Johnson wrote:
I also notice that I always get a warning during processing:
[java] Ambiguous rule match for /
[java] Matches both / on line 103 of
file:/c:/dpplus/dcbk/lib/docbook-xsl-ns/epub/docbook.xsl
[java] and / on line 404 of
davep wrote:
On 09/26/2011 08:05 PM, Keith Fahlgren wrote:
http://idpf.org/epub/30/spec/epub30-publications.html#confreq-rs-backward-epub2
It should process EPUB version 2 Publications as defined in[OPF2]
http://idpf.org/epub/30/spec/epub30-publications.html#refOPF2,[OPS2]
Hi Bob,
working with the latest Beta2 of the EPUB3 extensions I'm missing some
metadata in the resulting EPUB file (which are present if working with
the original EPUB2 XSL stylesheets):
bookinfo id=bookinfo
titleSome Title/title
subtitleSubtitle/subtitle
author
Keith Fahlgren wrote:
You can include the (somewhat deprecated) DCMES versions as well:
http://idpf.org/epub/30/spec/epub30-publications.html#sec-opf-dcmes-optional
This (duplication) will help older EPUB 2 reading systems (and should
probably be controlled with a parameter).
Actually I
Robert Nagle wrote, on 04.03.2011 18:50:
I certainly don't have Keith's programming skills, but I have been
compiling a lot of docbook-to-epub tips over the last year
http://www.imaginaryplanet.net/weblogs/idiotprogrammer/2010/11/ebookepub-production-secrets-tips-tricks/
One remark regarding
Lwam Berhane berh...@punkt.de wrote:
I want to join the german docbook mailinglist
http://placebo.hpi.uni-potsdam.de/mailman/listinfo/docbook. But my
browser could not find that page? does somebody knows why?
I'm not aware of any *active* German DocBook Mailinglist
(though I know there is
P
DeanNelson deannel...@aol.com schrieb:
All,
Wouldn't this be a real useful option in the var'lists? I know that in my own
conversions I needed to list in this manner but the only structure that would
look right is bold/underline the first term in the paragraph -which is wrong
on many
I'd like to see at least rudimentary support for
bookinfo|articleinfo
authorgroup
author
...
/author
/authorgroup
in OPF metadata of epub output.
This could be accomplished e.g. like
--- docbook.xsl.1.75.2 2009-06-16 04:59:35.0 +0200
+++ docbook.xsl 2009-10-23
What would be the preferred way to add meta information (like
dc:description,
dc:identifier, dc:title, or dc:language) given that the input file
might not contain the necessary information?
Should I simply override the opf template?
Michael
PS: Some TeX guys asked me about how to convert
Keith Fahlgren wrote, on 18.06.2009 17:41:
Sorry, I don't understand where this data would come from.
Some of this data might be part of the original Latex file (but might
not be converted to appropriate DocBook elements). As a last resort
there is always the possibility to add this data
Keith Fahlgren wrote, on 08.05.2009 22:40:
Please let me know if tonight's snapshot fixes this issue (or replace
your epub/bin/lib/docbook.rb with
https://docbook.svn.sourceforge.net/svnroot/docbook/trunk/xsl/epub/bin/lib/docbook.rb).
I now have a regression that captures this issue (I think).
Keith Fahlgren wrote, on 09.05.2009 19:08:
Thank you for the test case. I just committed another fix regression
to try to solve this issue flexibly. Please let me know if that
doesn't resolve it.
This works!
Without declared entities in the test file this version of docbook.rb
processes the
):
The new version complains like:
db-templ.xml:31: parser error : Entity 'copy' not defined
holdercopy; Michael Wiedmann/holder
about not declared entities, whereas earlier versions processed the same
file w/o such error (at least IIRC).
But:
If I look into the generated XHTML files in OEBPS
Stephane Bortzmeyer bortzme...@nic.fr wrote:
...
I do not find an option of dblatex to display the URL.
Nor do I.
At this time, the only solution I see is a bit of XSL to rewrite my
'ulink url=http://www.bortzmeyer.org/;blog/ulink' into 'ulink
url=http://www.bortzmeyer.org/;blog/ulink
Added 4 missing translation entries:
- Acknowledgements
- Authors
- References
- Translator
Michael
es.xml.gz
Description: GNU Zip compressed data
-
To unsubscribe, e-mail: docbook-apps-unsubscr...@lists.oasis-open.org
For
Keith Fahlgren wrote, on 13.04.2009 01:36:
Thanks for the bug report. This has been fixed in SVN revision 8407
a regression has been added with your test file. Please let me know if
that does not resolve your issue.
Verified!
With the latest snapshot all known remaining issues
I'm using the 2009-04-10 snapshot of the DocBook XSL stylesheets to take
advantage of the significant improvements I've seen in the repository
with respect to ePub metadata.
So far (almost) everything works great:
- attribute file-as of dc:creator works like expected
- dc:publisher and
* Mike M [EMAIL PROTECTED] wrote [030226 13:37]:
I did apt-get install sgml2x. I can't find runjade.
IIRC 'runjade' is only in the unstable 'sgml2x' package.
Michael
--
[EMAIL PROTECTED] http://www.miwie.org
[EMAIL PROTECTED]
* Bob Stayton [EMAIL PROTECTED] [030216 22:01]:
I think it is an oversight. Could you file a bug report
on http://sourceforge.net/projects/docbook/ ?
Done.
Michael
--
[EMAIL PROTECTED] http://www.miwie.org
[EMAIL PROTECTED]
I want to generate 'additional' HTML Code out of my DocBook XML file:
div class=navigatingbanner
p
a href=#section-id1Section Title #1/a
a href=#section-id1Section Title #1/a
a href=#section-id1Section Title #1/a
...
/div
'href' lines should be generated for every sect1 of my document.
Using
* Bob Stayton [EMAIL PROTECTED] [030214 09:30]:
a navigation banner at a fixed position in the browser
means you need to use an HTML Frameset. And a list of
sections would generally mean a table of contents, right?
Let me clarify what I want to achieve:
See e.g.
Please let me point you to the upcoming talk of Simon Pepping about his
work Docbook In ConTeXt, a ConTeXt XML mapping for DocBook documents
at the DANTE 2003 meeting (German TeX Users Group Meeting) at Bremen,
April, 2-4, 2003.
You'll find his abstract here:
* Bob Stayton [EMAIL PROTECTED] [030204 09:08]:
...
If you can't wait for a bug fix, you would need to copy
a template to your stylesheet customization layer
and fix it. The template is
xsl:template match=variablelist in html/lists.xsl.
You would need to change two things.
Thanks Bob,
For the following piece of DocBook XML the DocBook XSL 1.60.1
stylesheets don't generate HTML output for variablelistabstract
(the same for variablelistpara, variablelistsimpara).
Is there an easy way to change this?
variablelist
titleSome Title/title
abstractparavariablelist
A.R. (Tom) Peters wrote:
...
openjade:/usr/share/sgml/docbook/dtd/xml/3.1.7/docbook.cat:27:0:W: DTDDECL
catalog entries are not supported
openjade:/usr/share/sgml/docbook/stylesheet/dsssl/modular/catalog:28:0:W:
DTDDECL catalog entries are not supported
Use 'runjade' from the 'sgml2x'
* Norman Walsh [EMAIL PROTECTED] [030120 23:08]:
http://docbook.sourceforge.net/release/xsl/docbook-xsl-1.60.0.zip
Any particular reason why there is no 'docbook-xsl-1.60.0.tar.gz'?
Michael
--
[EMAIL PROTECTED] http://www.miwie.org
[EMAIL PROTECTED]
Using DocBook XML 4.2, DocBook XSL Stylesheets 1.59.2, and Saxon 6.5.2:
I want to limit the level of section's for which I get numbering in HTML
output (e.g. limit the level to 4).
Is there any parameter I can set in my customization layer or any
other solution?
Michael
--
[EMAIL PROTECTED]
* Hery [EMAIL PROTECTED] [021023 13:07]:
is there someone who tried direct conversion from an xml docbook to
latex file ?
There are several approaches to do this, but none of them really works
yet for 'real world' DocBook files:
- Ramon Casellas: DB2LaTeX XSL Transformation
Sebastian Bergmann wrote:
...
IMHO, a better translation would be
l:gentext key=nav-next text=Weiter/
l:gentext key=nav-next-sibling text=Schnell weiter/
I agree!
Michael
--
office: [EMAIL PROTECTED]
private: [EMAIL PROTECTED] http://www.miwie.org/
* Rachel Jensen [EMAIL PROTECTED] [020924 15:17]:
Hi. I am wondering how I turn the bookmark option off for my PDF output.
I am using the DSSSL stylesheets and OpenJade to convert the docbook markup
to PDF. I searched the archives and found several posts on formatting the
bookmarks but
* David Cramer [EMAIL PROTECTED] [020918 21:08]:
...
A related question regards xsl processors and sorting. I checked Saxon's
mailing list, but couldn't tell whether I could expect it to sort
correctly out-of-the-box or if we'd have to do a little work:
* Norman Walsh [EMAIL PROTECTED] [020904 07:03]:
/ Bob Stayton [EMAIL PROTECTED] was heard to say:
| So I think you should file a bug report on SourceForge
| to add a TOC to the article/appendix template, conditionally
| controlled by the TOC parameters. The article/appendix
| TOC could be
Does anybody have experience using 'docbook2X-0.7.0' to convert
DocBook XML documents to Texinfo?
My tests with even the simplest article's or book's are disappointing
because I get te weirdest errors in at least one of the processing steps.
Michael
--
[EMAIL PROTECTED]
I use the customized XSL of Sven Kitschke to generate HTML Help files
which was announced here some weeks ago
(http://lists.oasis-open.org/archives/docbook-apps/200205/msg00349.html).
This works great so far, but I cannot figure out how to generate a
real working 'Full-text search' in the
A few days ago (2002-02-11) Torsten Bronger announced his tbook system for
XML authoring in de.comp.text.tex which might IMO be interesting for some
people on this list:
quote
The tbook system for XML Authoring
Main Goals
An XML DTD that is sufficient for demanding, especially scientific,
Is there any reason why the solution for Chunk.xsl produce no valid
xhtml nor doctype as suggested in:
http://lists.oasis-open.org/archives/docbook-apps/200111/msg00343.html
is not incorporated in the DocBook XSL stylesheets?
In general it works like suggested and could make it much easier to
Marc Haber wrote:
...
Which toolchain is the recommended way for PDF creation from docbook
XML sources at this very moment?
For a working temporary solution - until the direct transformations from
FO to PDF work better - you should give 'htmldoc'
(http://www.easysw.com/htmldoc/) a try which
Using:
- DocBook XML 4.1.2
- DocBook XSL 1.51.1
- xsltproc (libxml2 2.4.22, libxslt 1.0.18)
In generating chunked HTML output I want to have TOCs up to the second
level and set the following in my XSL customization layer:
xsl:param name=generate.section.toc.level select=2/
But I get only a
Sven Kitschke wrote:
...
Some weeks ago i initially started to produce the attached
customization. I've rounded it up to some degree. It was tricky but it
works now. It may still contain bugs.
You possibly find it worth to include it in the stylesheets or to use
it as a starting point for
Using:
- DocBook XML 4.2.cr1-1
- DocBook XSL 1.50.1-EXP2
- xsltproc (libxml 20421, libxslt 10017, libexslt 708)
OR
saxon 6.4.4
Given the following minimal test file (processed with default
XSL stylesheets - not customized):
?xml version='1.0' encoding='ISO-8859-1'?
!DOCTYPE article PUBLIC
* Daniel Veillard [EMAIL PROTECTED] [020526 06:09]:
trying another XSLT processor to see if it's xsltproc fault's (I hope not,
but this is always possible ...)
Trying with saxon 6.4.4 gives correct result (no mis-sorted blocks)!
I will get the newest version of libmxl2 (2.4.21) and libxslt
Using:
- docbook 4.2.cr1
- docbook xsl stylesheets 1.50.0 (default html/docbook.xsl, no customization)
- xsltproc (libxml 20420, libxslt 10016, libexslt 707)
My source document contains a lot of indexterms and at the end of the
document an appendix which contains an empty index. The XSL
* Yann Dirson [EMAIL PROTECTED] [020516 23:14]:
I just started to investigate index generation. collateindex.pl appears to
collaborate only with the HTML stylesheets, and there is no similar
support-code in the print stylesheets. Does anyone know about tools/methods
that would be more
In a thread in 'de.comp.text.xml' concerning forced line-breaks in paras
Bernd Eckenfels posted two links [1,2] to postings which show that
Norm thought of implementing a processing instruction like ?line-break?
in the DSSSL/XSL stylesheets to support such forced line-breaks.
Is this still on
Yann Dirson wrote:
...
This may be off-topic, but HTMLDOC probably has most of the mechanics
implemented - just that DocBook is not (yet) in its list of output
formats (only PDF and postscript it seems). It's GPL'd software, so
any interested person could have a look at this.
I posted
Is there a way to reference the VERSION no. of the DSSSL stylesheets
somehow in my SGML sourcefiles?
I want to use something like
dsssl-version;
mainly for testing purposes and/or automatic inclusion in my source
files.
Michael
--
[EMAIL PROTECTED]
* Ian Castle [EMAIL PROTECTED] [020224 11:46]:
PassiveTeX takes this approach
see http://www.hcu.ox.ac.uk/TEI/Software/passivetex/
passivetex works on FO files, whereas using xmltex with a special
DocBook map file would create native LaTeX code and let (pdf)latex
typeset the document - a
* Kevin Dunn [EMAIL PROTECTED] [020224 13:17]:
...
docbook-dsssl-1.76 with the same tools produces PDF bookmarks for chapters and
sections, without the ability to collapse the sections into the chapters, i.e.
I can't reproduce this here.
Earlier this afternoon I tested the new DSSSL 1.76
* Ian Castle [EMAIL PROTECTED] [011221 07:29]:
...
The output from Jade with the tex-backend is a single file. The diagram,
if I am reading it correctly, implies that there is a separate output
for PDF and PS.
AFAIK the FreeBSD toolchain for creating documentation indeed creates
separate
camille wrote:
I just published at http://www.Linux-Mandrake.com/en/doc/project/ a
report written by a trainee. He worked on improving DSSSL stylesheets
and jadetex to fix many page setting problems we had on compiling
manuals from DocBook:
...
Has anyone successfully tested the use of the
camille wrote:
I just published at http://www.Linux-Mandrake.com/en/doc/project/ a
report written by a trainee. He worked on improving DSSSL stylesheets
and jadetex to fix many page setting problems we had on compiling
manuals from DocBook:
...
This is really great news!
Camille, do you
* G.L. Grobe [EMAIL PROTECTED] [011013 16:37]:
For doing something as simple as changing background color and the sizes of
titles on my main page, is there really no other way than to learn the
driver syntax and have to write a *.dsl file? I was browsing through all my
*.dsl files looking for
Dan York wrote:
[...]
Precisely! I have a Makefile set up for my user guide so that when I
simply type make all (on a Linux system), it:
- validates the SGML
- creates userguide.xml from userguide.sgml (changes the header)
- generates a PDF version (SGML/DSSSL)
- cd's into a
Given the line
ulink url=http://some.host.net/index.php3?para1=Xpara2=Y;xxx/ulink
(open)jade logically complains like:
general entity para2 not defined and no default entity
because of the ''.
The only workaround which comes to my mind is replacing '' with '%26'
or is there any other way
[ from docbook ML]
Rick Bronson wrote on 010617 23:00 -0700:
How to get rid of List of Tables in a book with chapters, each
chapter has one table.
I'd suggest to override the following piece of code from
print/dbparam.dsl in your customization layer:
(define ($generate-book-lot-list$)
Bob McIlvride wrote on 010614 09:32 -0400:
...
Could it be a jadetex issue? We're using version 1.63.
Using openjade and jadetex 3.5 produces *almost* correct PDF:
- the _ in the title text displays as expected,
- the entry in the TOC is correct too,
- but in the bookmarks the '_' just
If ndash; appears as index text like:
indextermprimaryfoo ndash; foo/primary/indexterm
this ndash; appears as (literally) #8211; in the printed index :-(
The corresponding TeX-code looks like:
...
{\def\fSize%
{9\p@}}foo\\#8211;,
\Node%
...
Any known workaround?
Michael
PS: jadetex 3.5
I wrote on 010612 22:37 +0200:
...
this ndash; appears as (literally) #8211; in the printed index :-(
...
Not only in print output, in HTML and RTF too!
So it looks like a stylesheet bug!
Michael
--
[EMAIL PROTECTED] http://www.miwie.org
[EMAIL PROTECTED]
Norman Walsh:
...
Sorry. Brain cramp. I misunderstood the question. Can you please submit
this as a feature request at http://sourceforge.net/projects/docbook/
Done, RequestID #431375
Michael
--
office: [EMAIL PROTECTED]
private: [EMAIL PROTECTED] http://www.miwie.org/
Applying plain DocBook DSSSL Stylesheets V1.70 on an article with
with qandasetqandadivqandaentry tags the first para of an
answer in 'print' output is rendered completely in 'bold'.
The second and following paras show up correct in 'normal'.
In older stylesheet versions (tested at least in
As suggested by a few members of this ML I make the SGML version of the
'DocBook DSSSL Stylesheet FAQ' available online too:
http://www.miwie.org/docbook-dsssl-faq.sgml
The HTML version is available at the usual place:
http://www.miwie.org/docbook-dsssl-faq.html
Beginning from Thursday I will
Carsten Wartmann pointed me to a problem in the generated HTML version
of the DocBook DSSSL FAQ:
All links for questions were pointing to #.QQ because of
missing id's for every question. After supplying every question with a
'id' attribute the links are correct (pointing e.g. to '#H1').
Is
Jirka Kosek wrote on 010420 12:35 +0200:
...
But in any case I would recommend you to switch DocBook qandaset way
of doing FAQs. DocBook is able to markup FAQs and both DSSSL and XSL
stylesheets supports this markup. It is like promoting BMW car sitting
in Volkswagen.
OK, I switched to
Sebastian Rahtz wrote on 010325 19:46 +0100:
so whats in the generated .tex file?
Though not a TeX-specialist I think the following code represents the
bookmark entries:
- bookinfo/title:
...
{1\p@}\def\LineSpacingFactor%
{0}}\def\HeadingText{%
}%
\endHeadPar{}\Seq%
{
...
- chapter/title
Norman Walsh wrote on 010325 08:52 -0500:
/ Michael Wiedmann [EMAIL PROTECTED] was heard to say:
| - I put a line like "pdfpagemode=UseOutlines" in my jadetex.cfg
I have no idea how this works. There's no obvious difference in the
way book titlepages are formulated, AFAICS. But
author-list-string should format authors in an authorgroup depending on
the chosen language, e.g.:
- English:
Firstname1 Surname1, Firstname2 Surname2, and Firstname3 Surname3
- German:
Firstname1 Surname1, Firstname2 Surname2 und Firstname3 Surname3
Norman Walsh wrote on 010321 11:36 -0800:
That's supposed to work. What mark are you trying to use?
Sorry, I should have looked closer at print/dblists.dsl. After your
reply I looked again at it and noticed "bullet", "box", "checkbox",
etc. and noticed that they work!
Thanks
Michael
--
DocBook SGML V4.1
DSSSL Stylesheets V1.62
jadetex 3.5
Is there an easy way to specify a "mark" attribute for an itemizedlist
which gets processed properly by the default print stylesheets?
Or are there any users who have done already some customization for the
stylesheets?
Michael
--
[EMAIL
Norman Walsh wrote on 010311 10:02 -0500:
/ Michael Wiedmann [EMAIL PROTECTED] was heard to say:
| Using DocBook 4.1 (SGML) and DSSSL stylesheets V1.62 I get unsatisfying
| output for affiliationaddressemail inside articleinfo:
Please send a small test document which demonstrates this bug
[ This must be a FAQ ]
Using DocBook 4.1 (SGML) and DSSSL stylesheets V1.62 I get unsatisfying
output for affiliationaddressemail inside articleinfo:
- in HTML output 10 leading and 8 trailing nbsp;
(at least I'd expect the same amount of nsbp;)
- in PDF output the email is not centered like
Norman Walsh wrote:
...
It works fine, for example, for DocBook XML V4.1.2, but when
attempting to compile the Website DTD, psgml chokes on the compiled
version of DocBook XML V4.1.2 (which it apparently tries to load when
it encounters a reference to it in website.dtd). I'm reasonably
82 matches
Mail list logo
