I have this rule in my XSL already:
<xsl:template match="title[starts-with(.,'Key Information for')]">
<xsl:copy>
<xsl:attribute name="outputclass">Key Information Long</xsl:attribute>
<xsl:apply-templates select="@* | node()"/>
</xsl:copy>
</xsl:template>
What it does: add outputclass='Key Information' to title tags that start with "Key Information for" during pre-processing in the DITA Open Toolkit.
What I need to do is add an XPath to a match rule that finds other titles that fit this pattern:
<title>Some title text (More title text)</title>
My xsl xpath needs to identify the titles that end in text contained in parentheses and add some new tagging (<ph aid:cstyle="parenthetical-subtitle">(More title text)</ph> to that piece of the title. When I add this rule, I also need to know how to integrate with the existing rule so it too gets the additional processing (and avoids an ambiguity conflict: both the existing rule and the new one could apply to the same instance of content).
I was imagining an edit involving choose/when logic with the ends-with() function, but I hit the wall on exactly how to do this, and it turns out the toolkit version I am using only supports xml 1.0. To illustrate --
If I have this input xml:
<section>
<title>Key Information for Our Test Process</title>
<p>Some stupid content.</p>
</section>
<section>
<title>Some Other Title</title>
<p>Other content.</p>
</section>
<section>
<title>Key Information for Testing (Stupid People Only)</title>
<p>Parentheses identify an audience which could really be anything.</p>
</section>
I want my edit to my rules to produce this output:
<section>
<title outputclass="Key Information Long">Key Information for Our Test Process</title>
<p>Some stupid content.</p>
</section>
<section>
<title>Some Other Title</title>
<p>Other content.</p>
</section>
<section>
<title outputclass="Key Information Long">Key Information for Testing <ph aid:cstyle="parenthetical-subtitle">(Stupid People Only)</ph></title>
<p>Parentheses identify an audience which could really be anything.</p>
</section>
If we can assume that the title element has no element children but solely plain text and you have an XSLT 2 or 3 processor as mentioning ends-with suggests you can match on the text node in the element and process it with xsl:analyze-string:
<xsl:template match="title[starts-with(.,'Key Information for')]">
<xsl:copy>
<xsl:attribute name="outputclass">Key Information Long</xsl:attribute>
<xsl:apply-templates select="@* | node()"/>
</xsl:copy>
</xsl:template>
<xsl:template match="title[starts-with(.,'Key Information for')]/text()">
<xsl:analyze-string select="." regex="\(.*\)$">
<xsl:matching-substring>
<ph aid:cstyle="parenthetical-subtitle">
<xsl:value-of select="."/>
</ph>
</xsl:matching-substring>
<xsl:non-matching-substring>
<xsl:value-of select="."/>
</xsl:non-matching-substring>
</xsl:analyze-string>
</xsl:template>