RE: [docbook-apps] [PDF] Bookmarks - add index groups

["Cramer, David W (David)" <dcramer@motive.com>]

Thanks for posting the details Jan. I've added a feature request linking back to this post. I think it would be a nice addition to the base xsls:
https://sourceforge.net/tracker/index.php?func=detail&aid=3057673&group_id=21935&atid=373750

David

-----Original Message-----
From: honyk [mailto:j.tosovsky@email.cz] 
Sent: Wednesday, September 01, 2010 1:10 PM
To: 'Jirka Kosek'; 'Bob Stayton'
Cc: docbook-apps@lists.oasis-open.org
Subject: RE: [docbook-apps] [PDF] Bookmarks - add index groups

Hello Guys,

thanks for your useful hints, it is working now!

Some details for future followers...

I've modified templates which match 'indexterm' in modes 'index-div-basic' and 'index-symbol-div' placing the id attribute on the block element (to create an anchor):

<fo:block id="id{$key}">

Into the bookmark processing code (mode="xep.outline") I've added another branching:
<xsl:when test="self::index">
  <rx:bookmark internal-destination="{$id}">
    <rx:bookmark-label>
       <xsl:value-of select="normalize-space($bookmark-label)"/>
    </rx:bookmark-label>
    <xsl:apply-templates select="." mode="xep.outline.index"/>
  </rx:bookmark>
</xsl:when>

which calls the following template

   <xsl:template match="index" mode="xep.outline.index">

      <xsl:param name="scope" select="(ancestor::book|/)[last()]"/>

      <xsl:variable name="role">
         <xsl:if test="$index.on.role != 0">
            <xsl:value-of select="@role"/>
         </xsl:if>
      </xsl:variable>

      <xsl:variable name="type">
         <xsl:if test="$index.on.type != 0">
            <xsl:value-of select="@type"/>
         </xsl:if>
      </xsl:variable>

      <xsl:variable name="terms"
         select="//indexterm
                        [count(.|key('letter',
                          translate(substring(&primary;, 1, 1),
                             &lowercase;,
                             &uppercase;))
                          [&scope;][1]) = 1
                          and not(@class = 'endofrange')]"/>

      <xsl:variable name="alphabetical"
         select="$terms[contains(concat(&lowercase;, &uppercase;),
                substring(&primary;, 1, 1))]"/>

      <xsl:for-each select="$alphabetical">
         <xsl:sort select="substring(&primary;, 1, 1)"/>
         <xsl:variable name="label">
            <xsl:call-template name="string.upper">
               <xsl:with-param name="string"
                  select="substring(&primary;, 1, 1)"/>
            </xsl:call-template>
         </xsl:variable>
         <rx:bookmark internal-destination="id{$label}">
            <rx:bookmark-label>
               <xsl:value-of select="$label"/>
            </rx:bookmark-label>
         </rx:bookmark>
      </xsl:for-each>

   </xsl:template>

The latter code requires placing special entities (/common/entities.ent) into the XSLT file:
<!DOCTYPE xsl:stylesheet [
<!ENTITY % common.entities SYSTEM "path to common entities ">
%common.entities;
]>

That's it.

Thanks once again.

Jan

> Bob Stayton wrote:
> 
> > The tricky part is determining which letters actually have entries in
> > the current index, so you don't create bookmark links to non-existant
> > index sections.  I don't have a quick solution for that one.
> 
> I think that code from autoidx.xsl can be reused, namely from
> generate-basic-template. For example following code will populate
> variable $alphabetical with exactly one indexterm for each letter.
> 
>   <xsl:variable name="terms"
>                 select="//indexterm
>                         [count(.|key('letter',
>                           translate(substring(&primary;, 1, 1),
>                              &lowercase;,
>                              &uppercase;))
>                           [&scope;][1]) = 1
>                           and not(@class = 'endofrange')]"/>
> 
>   <xsl:variable name="alphabetical"
>                 select="$terms[contains(concat(&lowercase;,
> &uppercase;),
>                                         substring(&primary;, 1, 1))]"/>
> 
> So in order to get just letters something like the following code could
> be used:
> 
> <xsl:for-each select="$alphabetical">
>   <rx:bookmark-label>
>     <xsl:value-of select="substring(&primary;, 1, 1)"/
>   </rx:bookmark-label>
> </xsl:for-each>
> 
> 				Jirka


---------------------------------------------------------------------
To unsubscribe, e-mail: docbook-apps-unsubscribe@lists.oasis-open.org
For additional commands, e-mail: docbook-apps-help@lists.oasis-open.org