Re: [docbook] Select chunk content from user.head.content template

["Bob Stayton" <bobs@sagehill.net>]

Hi Morten,
There is a utility template named "chunk" that can be used to test if a 
given node is a chunk element.  You could use it to check the first child 
section:

<xsl:variable name="chunk.children">
 <xsl:call-template name="chunk">
   <xsl:with-param name="node" select="section[1]"/>
 </xsl:call-template>
</xsl:variable>

The $chunk.children will have a value of 1 if that section is a chunk, or 0 
otherwise.  Then you can use that in the select statement:

<xsl:for-each select=".//indexterm[$chunk.children = 0] |
                    (indexterm 
|*[not(self::section)]//indexterm)[$chunk.children = 1]">

The select statement combines two parts that are mutually exclusive.
The first part selects all the indexterms, but qualifies them with a boolean 
expression to exclude all of them if $chunk.children = 1.
The second part selects indexterms that are either immediate children of the 
current section, or a descendant of any child element that is not a section. 
This part is qualified by the opposite boolean expression, so only one or 
the other part is used.

I didn't test this, so let me know if it doesn't work for you.  There might 
be a glitch that I'm not seeing.

Bob Stayton
Sagehill Enterprises
bobs@sagehill.net

--------------------------------------------------
From: "Olsen, Morten Engelhardt" <Morten_engelhardt.Olsen@atmel.com>
Sent: Tuesday, March 19, 2013 11:14 AM
To: <docbook@lists.oasis-open.org>
Subject: [docbook] Select chunk content from user.head.content template

> As you may or may not know, I have started on cleaning our internal 
> Microsoft Help Viewer 1.0 transformation up to a point where it is 
> releasable to the public. However, in this effort I have come across some 
> issues.
>
> Some background;
> The Microsoft HV1 format is the replacement for chm and hlp files. It is 
> based off xhtml with a set of header tags an some special magic. The set 
> of html and other content are zipped into a zip volume with the mshc 
> extension (Microsoft Help Collection or similar ☺). This file (called a 
> package) is accompanied by a manifest file with the extension msha. This 
> file can point to multiple packages, and this collection of packages is 
> called a book.
>
> Now to my first found issue;
> The header for a html file needs to contain a set of special meta tags. 
> This is what the help system uses to generate positional/relational 
> information, indexing and content description. Some of these tags needs to 
> contain information of the current chunk (e.g a set of <meta 
> name="Microsoft.Help.F1" content="myTopic" /> for all F1 content in the 
> current chunk). Currently we use indexterms with IDs to mark content for 
> F1 resolution, and the following template is called from 
> “user.head.content”:
>
> <xsl:template name="mshelp3.help.f1">
>      <xsl:for-each select="self::*//indexterm">
>            <xsl:if test="@id">
>                  <meta xmlns="http://www.w3.org/1999/xhtml"; 
> name="Microsoft.Help.F1">
>                        <xsl:attribute name="content">
>                             <xsl:value-of select="@id" />
>                        </xsl:attribute>
>                  </meta>
>            </xsl:if>
>      </xsl:for-each>
> </xsl:template>
>
> However, the “self::*//indexterm” matches all indexterms below self, but 
> with a chunk depth of e.g.4, the content of self may be chunked into 
> separate files. So how can I select the content of only the current chunk 
> from “user.head.content”? The current situation is that F1 ids are flowing 
> up in the chunking tree and usually hits on a higher level than were the 
> actual section exists.
>
>
> Morten Engelhardt Olsen
> Software Engineer / Tools Group / Atmel Corporation
> Tel: (+47) 930 94 593
>
> The information contained in this email message may be privileged, 
> confidential and/or protected from unauthorized disclosure. If you are not 
> the intended recipient, any dissemination, distribution or copying is 
> strictly prohibited. Please immediately notify the sender by reply if you 
> received this email in error. Thank you for your cooperation.