DITA Metrics Report: need @conkeyref support

[lief.erickson]

Hello--

The DITA Metrics report provides a lot good information, but it's incomplete. It does not provide reuse statistics for any @conkeyref. I'd like to add support for that. (Even better would be if oXygen had default support for it. I know others would like to see it, too. http://www.oxygenxml.com/pipermail/oxyg ... 04799.html)

I have been able to enhance the files in frameworks\dita\report to show me how many @conkeyref are used and what elements they appear in. What I can't figure out is how to "gather all the key definitions that form the keyspace - these can be placed in a variable that is used then to lookup for the corresponding node in the resolve.xsl stylesheet when we match on an element having a conkeyref on it" as George Bina wrote in the above post from October 2013. I'm guessing that if this were easy oXygen would already have this support.

Is there template or stylesheet that someone could point me to that might give me some hints on how to do this?

-Lief

[george]

Hi Lief,

Well, it is not that difficult... I just implemented support for determining the key space and handling topicref/@keyref and @conkeyref and this will be generally available in the next oXygen release. If you need this now please contact us directly at support@oxygenxml.com and I will send you the updated XSLT scripts.

Best Regards,
George

[lief.erickson]

A message has been sent. Thanks, George.

[Frank Ralf]

Hi George,

Is this feature already implemented in oXygen 16?

Best regards,
Frank

[george]

Hi Frank,

I think I implemented this in 18.0 or 17.1 - I do not remember exactly - just update to version 18 and you should have it there.

Regards,
George

[Frank Ralf]

Hi George,

Thanks for this information. Our client is still on oXygen 16.1 but I will have a look in oXygen 17 or 18.

Kind regards,
Frank

[Frank Ralf]

It's there in oXygen 17.1, in Oxygen XML Editor 17\frameworks\dita\report\report.xsl

Code: Select all

 

 <!-- CONKEYREFs -->

    <xsl:template match="*[@conkeyref]" mode="resolve-base resolve">

        <xsl:param name="keyspace" tunnel="yes"/>

        <xsl:variable name="key" select="if (contains(@conkeyref, '/')) then substring-before(@conkeyref, '/') else @conkeyref"/>

        <xsl:variable name="elementID" select="substring-after(@conkeyref, '/')"/>

        <xsl:variable name="keydef" select="$keyspace/oxyd:keyspace/oxyd:key[@value=$key]"/>

        <xsl:variable name="topicFile" select="$keydef/*[1]/document(@href, .)"/>

        <oxyd:conref element="{name()}" key="{$key}">

            <xsl:attribute name="xml:base" select="document-uri($topicFile)"/>

            <xsl:apply-templates select="if ($elementID = '') then $topicFile/* else $topicFile//*[@id=$elementID]" mode="#current"/>

        </oxyd:conref>

    </xsl:template>

Frank

[Frank Ralf]

EDIT:

The above code resides in Oxygen XML Editor 17\frameworks\dita\report\modules\resolve.xsl

Frank

[Frank Ralf]

I've just created a report from the same DITA map with the oXygen 16.1 and oXygen 17.1 versions of the report script. I don't quite understand the results because the oXygen 17 version reports more reused elements (because of the @conkeyref'ed elements) and words but a lover reuse percentage. How can that be?

Frank

Content reuse - oXygen 16 - without @conkeyref
Total reused words (words in conref content) 52,328. Content reuse percentage (words) is 24.74%.
Total reused elements (elements in conref content) 11,464. Elements reuse percentage is 9.13%.
Total content reference elements 27.

Content reuse - oXygen 17 - with @conkeyref
Total reused words (words in conref content) 52,511. Content reuse percentage (words) is 22.19%.
Total reused elements (elements in conref content) 11,784. Elements reuse percentage is 7.85%.
Total content reference elements 334.

[george]

Hi Frank,

You need to check the "Total words" value and the "Total topic elements" - these are the values that represent 100% of words and topic elements, respectively.

Best Regards,
George

[Frank Ralf]

Hi George,

Thanks for the clarification.

Kind regards,
Frank