OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

Mail Index message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: docbook-apps-digest Digest #444

DOCBOOK-APPS: Problems with xsl stylesheets (xalan)
DOCBOOK-APPS: DB xsl V1.60.1 - email formatting
DOCBOOK-APPS: Re: srcdir/objdir issues and including variable
 usability issues with xsltproc
Re: DOCBOOK-APPS: how-to make an index using xsl 1.60
DOCBOOK-APPS: Table titles
Re: DOCBOOK-APPS: Installing and using docbook2X (debian)
DOCBOOK-APPS: Ant and XMLSpy
RE: DOCBOOK-APPS: Ant and XMLSpy
Re: DOCBOOK-APPS: Problems with xsl stylesheets (xalan)
Re: DOCBOOK-APPS: Table titles
Re: DOCBOOK-APPS: DB xsl V1.60.1 - email formatting
Re: DOCBOOK-APPS: How to select <imageobject> alternative in <med
 iaobject>?
Re: DOCBOOK-APPS: How to select <imageobject> alternative in <med
 iaobject>?
Re: DOCBOOK-APPS: How to select <imageobject> alternative in <med
 iaobject>?
DOCBOOK-APPS: FOP: "No ImageReader" when using PDF as vector format
DOCBOOK-APPS: generating PDF using FOP?
DOCBOOK-APPS: qnaentry in div, not html wrapper
DOCBOOK-APPS: XIncludes
Re: DOCBOOK-APPS: qnaentry in div, not html wrapper
Re: DOCBOOK-APPS: qnaentry in div, not html wrapper
Re: DOCBOOK-APPS: qnaentry in div, not html wrapper
DOCBOOK-APPS: <beginpage /> fails with fop xsl 1.60.1
DOCBOOK-APPS: Sectioning Problem

----------------------------------------------------------------
To subscribe or unsubscribe from this elist use the subscription
manager: <http://lists.oasis-open.org/ob/adm.pl>


Hi,

I've run into major problems converting an xml docbook document
into (x)html.

I use:
- article.xml in valid docbook 4.2 format
- java: JAXP=1.1, SAX=2.0, xalan2=2.2.D11, DOM=2.0
- docbook-xsl-1.60.1

1) converting  article to html using html/chunk.xsl works.

2) converting  article to html using xhtml/chunk.xsl produces
   weired symbols in the output document.    
e.g.
- "Â|" in almost all navigation menues. 
   (having an <url> link in <sec1> removes suppresses this charakter
- "â~@~]"
   before and after <quote> sections

How can i remove those ?

Ciao

Jens Skripczynski
-- 
E-Mail: skripczynski(at)mail2003(dot)skripczynski(dot)de

The light at the end of the tunnel 
is just the head light of the oncoming train.





Using latest DB xsl v 1.60.1, 4.2 docbookx.dtd,
the email formatting seems to be broken,
esp. when compared to 1.59.1.

It seems to keep email with surname on same line
for FO (pdf) output:

<article> <articleinfo> <title> X </title>
<author> <firstname> A Firstname </firstname>
  <surname> MySurName </surname>
  <affiliation> <address> 
   <email>email_in_wrong_place@nowhere.com</email>
   </address></affiliation>
</author>
</articleinfo>
</article>

Fo Output w 1.60.1 xsl, DB4.2 dtd: 
A Firstname
MySurname<email_in_wrong_place@nowhere.com>

Is this a bug? Is there a recommended work-around?
thanks-
K

__________________________________________________
Do you Yahoo!?
Yahoo! Tax Center - forms, calculators, tips, more
http://taxes.yahoo.com/



>>>>> "JL" == John Levon <levon@movementarian.org> writes:

JL> I've found this stuff to be entirely impenetrable. Is there some
JL> good reason why a sensible, simple include and variable
JL> substitution system isn't available ?

[...]

JL> <xsl:param name="srcdir"/> <xi:include
JL> href="$srcdir/xsl/xhtml-non.xsl"
JL> xmlns:xi="http://www.w3.org/2001/XInclude" />

JL> So my questions are :

JL> 1) (general point) why is this so hard ? Where is xsltproc -I and
JL> xsltproc -D ?  2) (specific point) how do I do it ?

JL> Examples in particular are appreciated. In case it's relevant :
JL> [moz@lambent shared]$ rpm -qf `which xsltproc` libxslt-1.0.19-1
JL> and xsl-stylesheets-1.50.0-3

JL> My attempts to find explanations and help generally end up
JL> pointing to the standards, which is not far off line noise for me.

The answer for the DocBook stylesheet location identificaton is to use
XML catalogs!  No need to muck around with autconf for this.

See:

 http://www.sagehill.net/xml/docbookxsl/Catalogs.html

and more specifically to locate XSL stylesheets:

 http://www.sagehill.net/xml/docbookxsl/Catalogs.html#d0e1733

By default xsltproc will look for catalogs in /etc/xml/catlog, so if
you set that up to point to right place(s) for the installed DocBook
files, your user's won't need to modify their environment at all.

Alex



Jens Skripczynski wrote:

> How can i automatically construct an index using '<indexterm>'  with xalan ?

Put empty <index/> element in a place where index should occur.

-- 
-----------------------------------------------------------------
  Jirka Kosek  	                     
  e-mail: jirka@kosek.cz
  http://www.kosek.cz



How do I turn off table titles (like "Table 2.1"). I want to use section
titles instead, and don't need the table numbering or title labels.

Brian




I have discovered workarounds for the issue that I was having earlier
installing docbook2X, but now I have a different one:

make[1]: Entering directory `/root/docbook2X-0.7.0/doc'
/usr/bin/perl -I../perl ../perl/docbook2manxml --header-3="`date "+%d %B %Y"`"
\
                  --refs-file=docbook2X.refs \
                  docbook2X.xml >docbook2X.mxml 2>/dev/null \
|| /usr/bin/perl -I../perl ../perl/docbook2manxml --header-3="`date "+%d %B
%Y"`" \
                     --refs-file=docbook2X.refs \
                     docbook2X.xml >docbook2X.mxml

read error at /usr/lib/perl5/XML/Parser/Expat.pm line 469.
 at line 29, column 21, byte 785
Handler couldn't resolve external entity at line 29, column 21, byte 785
--snip--

Line 29 looks like this:
<releaseinfo>Version &db2x_version;</releaseinfo>

It looks like my version of XML::Parser (2.30) does things differently than
XML::Parser 2.27 that is listed on the docbook2x.sf.net site. 
Does anyone know how to resolve this?

__________________________________________________
Do you Yahoo!?
Yahoo! Web Hosting - establish your business online
http://webhosting.yahoo.com



If anyone is interested, I have found XMLSpy (v5r3) and Ant to work well 
together. I searched this list for "Ant" and found nothing.

I pass the build.xml file in as if it were the stylesheet :)

Here is the command line I use for the "External XSL Transformation 
program:" in the Options window:

C:\programs\Ant\jakarta-ant-1.5.1\bin\ant.bat -Dsource.file=%1 -f %3 
-Dout.slot=%2 -Dstyle.file=%3

Ant does the rest (I haven't tried to figure out why I needed the absolute 
path).

This Atom Ant (Super Spy?) combination allows me to use different XSLT 
processors through XMLSpy without opening the Options window each time I 
want to change. Also, I intend it to replace Make when I am using the 
WebSite doctype. Mr. Walsh said something about needing Make for sites with 
source in more than one directory.

(I couldn't transform DocBook with either the native XMLSpy XSLT processor 
or MSXML, btw)

I have used this combo with Saxon, Jade, and Xalan. TeX causes it to freeze 
though. TeX likes to ask questions, so I use Ant to build the FO and control 
TeX from the command prompt.

I was pretty happy to find this combo and am hoping someone else will also 
find this info useful.

Now, if I could only get paid to do this stuff! ;)

p.s. I do have a question. What are the "digest" email lists for, like 
docbook-apps-digest?

sds

_________________________________________________________________
The new MSN 8: advanced junk mail protection and 2 months FREE*  
http://join.msn.com/?page=features/junkmail




You must have searched via the oasis archives [1] where "Words must be
at least 4 letters or digits long to be indexed"[2]. Try the
sources.redhat.com archive [3] and you'll get some hits.

[1] http://lists.oasis-open.org/archives/docbook-apps/
[2] http://lists.oasis-open.org/elists/search-help.shtml
[3] http://sources.redhat.com/ml/docbook-apps/

David

-----Original Message-----
From: Scott Shepard [mailto:scottdavidshepard@hotmail.com] 

<snip/> 
together. I searched this list for "Ant" and found nothing.
<snip/>




On Sun, Mar 09, 2003 at 07:46:25AM +0100, Jens Skripczynski wrote:
> Hi,
> 
> I've run into major problems converting an xml docbook document
> into (x)html.
> 
> I use:
> - article.xml in valid docbook 4.2 format
> - java: JAXP=1.1, SAX=2.0, xalan2=2.2.D11, DOM=2.0
> - docbook-xsl-1.60.1
> 
> 1) converting  article to html using html/chunk.xsl works.
> 
> 2) converting  article to html using xhtml/chunk.xsl produces
>    weired symbols in the output document.    
> e.g.
> - "B|" in almost all navigation menues. 
>    (having an <url> link in <sec1> removes suppresses this charakter
> - "b~@~]"
>    before and after <quote> sections
> 
> How can i remove those ?

Those "weird" characters are the same characters you
get with html/chunk.xsl, except they are in UTF-8
encoding instead of character entities.
You'll notice that the top of each output file says

<?xml version="1.0" encoding="UTF-8"?>

so the characters are actually correct and should
display ok in most browsers.

If you truly don't like them, you need to change
the output encoding.  The default output encoding
for the xhtml stylesheets is UTF-8.  Since Xalan doesn't
accept an option to set the output encoding the way Saxon
or xsltproc (with exslt) does, you will have to create an
XSL customization file that sets:

<xsl:output method="xml" encoding="iso-8859-1"/>

Then those special characters will like be output
as character entities instead.

-- 

Bob Stayton                                 400 Encinal Street
Publications Architect                      Santa Cruz, CA  95060
Technical Publications                      voice: (831) 427-7796
The SCO Group                               fax:   (831) 429-1887
                                            email: bobs@sco.com



On Mon, Mar 10, 2003 at 05:18:32PM -0500, Brian Burridge wrote:
> How do I turn off table titles (like "Table 2.1"). I want to use section
> titles instead, and don't need the table numbering or title labels.

If you don't want table titles, then you should use
<informaltable> instead of <table>.  The only difference
is the <title> element.

-- 

Bob Stayton                                 400 Encinal Street
Publications Architect                      Santa Cruz, CA  95060
Technical Publications                      voice: (831) 427-7796
The SCO Group                               fax:   (831) 429-1887
                                            email: bobs@sco.com



On Sat, Mar 08, 2003 at 01:48:35PM -0800, Kevin A. Chin wrote:
> 
> Using latest DB xsl v 1.60.1, 4.2 docbookx.dtd,
> the email formatting seems to be broken,
> esp. when compared to 1.59.1.
> 
> It seems to keep email with surname on same line
> for FO (pdf) output:
> 
> <article> <articleinfo> <title> X </title>
> <author> <firstname> A Firstname </firstname>
>   <surname> MySurName </surname>
>   <affiliation> <address> 
>    <email>email_in_wrong_place@nowhere.com</email>
>    </address></affiliation>
> </author>
> </articleinfo>
> </article>
> 
> Fo Output w 1.60.1 xsl, DB4.2 dtd: 
> A Firstname
> MySurname<email_in_wrong_place@nowhere.com>
> 
> Is this a bug? Is there a recommended work-around?

I'm not sure what you are seeing.  I think your email
line breaks changed.  The output from
1.59.1 is something like:

                      A Firstname MySurName


                <email_in_wrong_place@nowhere.com>

I get a lot of vertical space with 1.59.1.
The output from 1.60.1 is:

      A Firstname MySurName <email_in_wrong_place@nowhere.com>

I think this was an intentional change in the stylesheet to
accomodate a document with multiple authors and emails.

If you don't like it, then copy the old template for
author from fo/titlepage.xsl to a customization
layer for 1.60.1

<xsl:template match="author" mode="titlepage.mode">
  <fo:block>
    <xsl:call-template name="anchor"/>
    <xsl:call-template name="person.name"/>
    <xsl:apply-templates select="affiliation" mode="titlepage.mode"/>
  </fo:block>
</xsl:template>


-- 

Bob Stayton                                 400 Encinal Street
Publications Architect                      Santa Cruz, CA  95060
Technical Publications                      voice: (831) 427-7796
The SCO Group                               fax:   (831) 429-1887
                                            email: bobs@sco.com



I'm using the 1.60.1 distro of the xsl stylesheets and have been
experimenting with alternative imageobjects in a mediaobject for html and
fo.

In short it doesn't work.  I have searched through the stylesheets and have
found that though the parameter "use.role.for.mediaobject" is documented and
is included in files such as param.xsl, it doesn't appear at all in any of
the stylesheets proper such as graphics.xsl!

I am wondering if this feature has been backed out or been delayed for a
future distro or what?

It's a very good feature, but where can I get it?!

nik


----- Original Message -----
From: "Bob Stayton" <bobs@caldera.com>
To: "Jeff Beal" <jeff.beal@ansys.com>; <docbook-apps@lists.oasis-open.org>
Sent: 06 December 2002 23:51
Subject: Re: DOCBOOK-APPS: How to select <imageobject> alternative in <med
iaobject>?


> Based on the feedback I got about this, I went ahead and
> checked in changes to the XSL stylesheets to support the
> use of the role attribute to select one of several
> possible imageobjects.
>
> This will let you put several imageobjects (or other
> *objects) in a mediaobject container.  If one object
> has a role="html" attribute, then that one will be used
> for HTML (and XHTML) output.  If there are more than one object
> with that role, then the selection process proceeds as
> before but just with those objects.  If none of
> the objects has that role, then it falls back to picking
> the first suitable object.
>
> For fo output, an object with role="fo" will be used in
> a similar manner.
>
> Per Jeff Beal's suggestion, I added a
> 'preferred.mediaobject.role' parameter to let you
> set your own name instead of 'html' or 'fo'.
> If that parameter is not blank and there is an object
> with that role value, then it is used.  If that parameter is
> not blank and there is no object with that role, then
> it falls back to trying to select 'html' or 'fo' objects.
> Otherwise it falls back to picking the first suitable
> object.
>
> Per Jirka's suggestion, I added a 'use.role.for.mediaobject'
> parameter switch to turn off the whole process of
> considering role in selecting the object.  Then it takes
> the first suitable object.  This can be used with
> profiling on the role attribute to select the object in the
> profiling stage.
>
> These changes will be in the next XSL distribution.
>
> Bob Stayton                                 400 Encinal Street
> Publications Architect                      Santa Cruz, CA  95060
> Technical Publications                      voice: (831) 427-7796
> The SCO Group                               fax:   (831) 429-1887
>                                             email: bobs@sco.com
>
>
> On Tue, Nov 26, 2002 at 09:17:57AM -0500, Jeff Beal wrote:
> > I agree that role is the best attribute for the author specifying which
> > object to use for which output.  I don't think that simply 'html' and
'fo'
> > will be able to satisfy every condition, though.  For example, we
support a
> > wide variety of platforms and many of our UNIX users use an X Windows
> > emulator to run our application from their Windows desktop.  Using the
> > audioobject would not work, simply because sound doesn't work in that
> > configuration.  However, we may want to use an audioobject for our
Windows
> > users.  There are probably several other examples.  (e.g. different
image
> > support between XEP and FOP)
> >
> > I think that if you add a "preferred.mediaobject.role" parameter, we can
> > modify the selection algorithm as follows:
> >
> > 1) Look for an object with "@role = $preferred.mediaobject.role"
> > 2) Look for an object where the role matches the output (html or fo)
> > 3) Pick the first object with either role="all" or no role attribute
> > 4) Look for a textobject
> >
> > Jeff Beal
> >
> > -----Original Message-----
> > From: Bob Stayton [mailto:bobs@caldera.com]
> > Sent: Monday, November 25, 2002 11:25 PM
> > To: docbook-apps@lists.oasis-open.org
> > Subject: Re: DOCBOOK-APPS: How to select <imageobject> alternative in
> > <mediaobject>?
> >
> >
> > The mediaobject element can contain more than one
> > imageobject element, with the intent being to use
> > one from the set for a given output.  But
> > there currently isn't an easy way of designating this
> > particular imageobject for HTML output and that
> > one for FO output.
> >
> > I suggested in a previous message that role="html" or
> > role="fo" in each imageobject could be used, with a
> > stylesheet customization to act on the role value.  I'm
> > wondering if this method would be appropriate as a standard
> > feature of the stylesheets? That is, should the stylesheets
> > designate the role attribute as the method to select
> > which imageobject to use?
> >
> > Let me be clear that this is not a change to the DTD, but
> > just a convention to be used with Norm's XSL stylesheets.
> >
> > There doesn't seem to be another available attribute that
> > would be appropriate.  And I'm guessing that any profiling
> > would probably be done on the role attribute in
> > mediaobject, rather than the individual imageobjects within
> > it.
> >
> > But I'd like to hear what others think.
> > Would this be useful?  Is there a better way to do it?
> >
> >
> > Bob Stayton                                 400 Encinal Street
> > Publications Architect                      Santa Cruz, CA  95060
> > Technical Publications                      voice: (831) 427-7796
> > The SCO Group                               fax:   (831) 429-1887
> >                                             email: bobs@sco.com
>
> --
>
> Bob Stayton                                 400 Encinal Street
> Publications Architect                      Santa Cruz, CA  95060
> Technical Publications                      voice: (831) 427-7796
> The SCO Group                               fax:   (831) 429-1887
>                                             email: bobs@sco.com




Nicholas Atkinson <nik@casawana.com>:
> I'm using the 1.60.1 distro of the xsl stylesheets and have been
> experimenting with alternative imageobjects in a mediaobject for html and
> fo.
>
> In short it doesn't work.

In short it does work :-)
I used it sucesfully with 1.60.1 with xsltproc and FOP.
Have you set both 'preferred.mediaobject.role' and
'use.role.for.mediaobject' ?

ABX



Yes, you're quite right.

It was a problem in my customization. (As usual!)

thanks

nik

----- Original Message -----
From: "ABX" <abx@abx.art.pl>
To: <docbook-apps@lists.oasis-open.org>; <nik@casawana.com>
Sent: 11 March 2003 12:38
Subject: Re: DOCBOOK-APPS: How to select <imageobject> alternative in <med
iaobject>?


Nicholas Atkinson <nik@casawana.com>:
> I'm using the 1.60.1 distro of the xsl stylesheets and have been
> experimenting with alternative imageobjects in a mediaobject for html and
> fo.
>
> In short it doesn't work.

In short it does work :-)
I used it sucesfully with 1.60.1 with xsltproc and FOP.
Have you set both 'preferred.mediaobject.role' and
'use.role.for.mediaobject' ?

ABX





Hello,

Bob Stayton writes in "Using the XSLT Sheets" that
FOP can handle PDF as a vector format:

Table 16.6. FO processor supported graphics formats
FO Processor Bitmap formats Vector formats
FOP PNG JPEG SVG PDF


But I get the following error from FOP 0.20.5rc2

[ERROR] Error while creating area : No ImageReader for this type of 
image (file:/KM/usr/ziegler/LEDA/Tutorial/pdf/Pictures/ArrayOfChar.pdf)


My DocBook source is the following:

   <mediaobject>
     <imageobject  role="html">
       <imagedata  align="center" format="PNG" 
fileref="Pictures/ArrayOfChar.png"/>
   </imageobject>

     <imageobject  role="fo">
       <imagedata  align="center" format="PDF" 
fileref="Pictures/ArrayOfChar.pdf"/>
     </imageobject>
   </mediaobject>


Can someone please help me? I've already browsed through the mailinglist 
archive and the bug reports of FOP, but I haven't found anything.


I'm using java version "1.3.1", DocBook Stylesheets 1.60.1


Greetings,
Joachim









  i'm going to see what kind of success i have using FOP
to generate PDF from my docbook files.  i've written a short,
several-page DB file for beginners that shows examples of
a pile of the simpler DB markup, such as simple inline
stuff, lists, programlistings and so on.

  i ran this through FOP already and, oh dear, there are
numerous weird renderings, although i was warned on the
FOP list that this might happen.

  anyone out there rendering their DB files using FOP,
and can fill me in on what they've had to hack so far?
or good URLs to help me start fixing some of these
issues?

  and if anyone wants to play along, i can email them
my sample DB file showing the various markup elements
and they can see what they get out of it.

rday

--

Robert P. J. Day, RHCE
Eno River Technologies, Inc.
Unix, Linux and Open Source training




I'm playing with docbook via a servlet,
with queries into mySQL database.
That means I may get n results.
   Each row would refer to a single qnaentry.

I'd like to wrap them all in a single html document,
each entry in some wrapper.

Has anyone looked at this class of customisation please?

regards DaveP





I use xerces/xalan/fop and I believe Xerces doesn't handle XIncludes at this
time.  What would you recommend for handling XIncludes?

I looked at http://xincluder.sourceforge.net, would this be the way to go?




On Tue, Mar 11, 2003 at 04:34:43PM +0000, Dave Pawson wrote:
> I'm playing with docbook via a servlet,
> with queries into mySQL database.
> That means I may get n results.
>    Each row would refer to a single qnaentry.
> 
> I'd like to wrap them all in a single html document,
> each entry in some wrapper.
> 
> Has anyone looked at this class of customisation please?

I'm not completely clear on what you are trying to do.
Is your servlet converting the data to DocBook elements?
Couldn't you create a DocBook document with
root element <qandaset> and each entry in
a <qandaentry>?  If you give it a title and
turn on the qandaset toc, then you wouldn't need any
stylesheet customization.

If you want each entry (both Q and A) in a <div> wrapper,
then something like this should work:

<xsl:template match="qandaentry">
  <div class="qandaentry">
    <xsl:apply-templates/>
  </div>
</xsl:template>


-- 

Bob Stayton                                 400 Encinal Street
Publications Architect                      Santa Cruz, CA  95060
Technical Publications                      voice: (831) 427-7796
The SCO Group                               fax:   (831) 429-1887
                                            email: bobs@sco.com



At 10:55 11/03/2003 -0800, Bob Stayton wrote:
>On Tue, Mar 11, 2003 at 04:34:43PM +0000, Dave Pawson wrote:
> > I'm playing with docbook via a servlet,
> > with queries into mySQL database.
> > That means I may get n results.
> >    Each row would refer to a single qnaentry.
> >
> > I'd like to wrap them all in a single html document,
> > each entry in some wrapper.
> >
> > Has anyone looked at this class of customisation please?
>
>I'm not completely clear on what you are trying to do.
>Is your servlet converting the data to DocBook elements?

1. Take a keyword from a user
2. Use that to look up in MySQL.
3. Returns the filename (xml docbook)
4. Use XSLT and Norms stylesheets to convert to html
5. Present that at the servlet as the 'result'.




>Couldn't you create a DocBook document with
>root element <qandaset> and each entry in
>a <qandaentry>?  If you give it a title and
>turn on the qandaset toc, then you wouldn't need any
>stylesheet customization.

Each 'answer' will be one qandaentry,
and since I may want to display n of them,
I need one html wrapper,
then n 'divs' or something, one per qandaentry.




>If you want each entry (both Q and A) in a <div> wrapper,
>then something like this should work:
>
><xsl:template match="qandaentry">
>   <div class="qandaentry">
>     <xsl:apply-templates/>
>   </div>
></xsl:template>

So......
    Provide my own template for the outer,
    then xsl:include ... what Bob?

    Surely not docbook.xsl?

regards DaveP






On Tue, Mar 11, 2003 at 07:48:20PM +0000, Dave Pawson wrote:
> At 10:55 11/03/2003 -0800, Bob Stayton wrote:
> >On Tue, Mar 11, 2003 at 04:34:43PM +0000, Dave Pawson wrote:
> > > I'm playing with docbook via a servlet,
> > > with queries into mySQL database.
> > > That means I may get n results.
> > >    Each row would refer to a single qnaentry.
> > >
> > > I'd like to wrap them all in a single html document,
> > > each entry in some wrapper.
> > >
> > > Has anyone looked at this class of customisation please?
> >
> >I'm not completely clear on what you are trying to do.
> >Is your servlet converting the data to DocBook elements?
> 
> 1. Take a keyword from a user
> 2. Use that to look up in MySQL.
> 3. Returns the filename (xml docbook)
> 4. Use XSLT and Norms stylesheets to convert to html
> 5. Present that at the servlet as the 'result'.
> 
> 
> 
> 
> >Couldn't you create a DocBook document with
> >root element <qandaset> and each entry in
> >a <qandaentry>?  If you give it a title and
> >turn on the qandaset toc, then you wouldn't need any
> >stylesheet customization.
> 
> Each 'answer' will be one qandaentry,
> and since I may want to display n of them,
> I need one html wrapper,
> then n 'divs' or something, one per qandaentry.
> 
> 
> 
> 
> >If you want each entry (both Q and A) in a <div> wrapper,
> >then something like this should work:
> >
> ><xsl:template match="qandaentry">
> >   <div class="qandaentry">
> >     <xsl:apply-templates/>
> >   </div>
> ></xsl:template>
> 
> So......
>     Provide my own template for the outer,
>     then xsl:include ... what Bob?
> 
>     Surely not docbook.xsl?
 
> 1. Take a keyword from a user
> 2. Use that to look up in MySQL.
> 3. Returns the filename (xml docbook)
> 4. Use XSLT and Norms stylesheets to convert to html
> 5. Present that at the servlet as the 'result'.

Doesn't "Surely not docbook.xsl" contradict #4?
Why not docbook.xsl?

Anyway, what I was suggesting is that the servlet do this:

3.1  Create a DocBook wrapper document:

<?xml version="1.0"?>
<!DOCTYPE qandaset PUBLIC blah balh>
<qandaset>
<?dbhtml toc="1" ?>
<title>Answers to your question</title>
<xi:include href="first.xml"/>
<xi:include href="second.xml"/>
<xi:include href="third.xml"/>
</qandaset>

Then step 4 processes that to html using xsltproc
with the --xinclude option and docbook.xsl.
Why not docbook.xsl?  Too big, too slow?

-- 

Bob Stayton                                 400 Encinal Street
Publications Architect                      Santa Cruz, CA  95060
Technical Publications                      voice: (831) 427-7796
The SCO Group                               fax:   (831) 429-1887
                                            email: bobs@sco.com



hi again,

I've written an article, that becomes longer and longer and i started
wondering if i could force page-break.

The docbook lists the <beginpage /> option, but it seems, that
the xsl stylesheet does not know about this. (I have a red
<beginpage></beginpage> in the pdf output file).

The xsl Fop deffinition does:
7.29.17 "page-break-before"
http://www.w3.org/TR/2001/REC-xsl-20011015/slice7.html#page-break-before

I'm not too familiar with docbook (and do know nothing about FOP), so is 
this a bug, or just missing in the stylesheet.

If it is missing, where do i have to add this as a customization ?

How about an extension (like the one listed in the FOP specs) 
 for <begin-page />? 

e.g.:
break=auto | always | avoid | left | right | inherit

with the default to break="always" as i do understand the deffinition
in the documentation.


P.S.: I think it would be good to add to the index section on
      "http://docbook.org/tdg/en/html/index.html", 
      that a simple <index /> tells processor where to put the
      index. I searched for this hint some time.
      thx to Jirka.

Ciao

Jens Skripczynski
-- 
E-Mail: skripczynski(at)mail2003(dot)skripczynski(dot)de

War doesn't determine who is right,  war determines who is left
                         (Chinese Proverb)




I changed my stylesheet to page break for level 1 sections (and
chapters), which has worked great. However, now I'm doing my preface,
and the preface has some sections. Since they are top level sections,
they are page breaking before each one, which is not what I want. I
really want to be able to use level 3 sections in my preface without
level 1 or 2, which of course is not possible.

I tried using simplesect, but apparently it still functions as a level 1
section. Is there anyway I can make simplesect function separately from
a level 1 section? So I could have level 1's auto number and page break,
but have simple sects not number, not be included in the TOC and not
page break?

Brian

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Powered by ezmlm-idx