Hi Folks,

Consider this element:

<style>
   blah, blah
   <![CDATA[
      <!--
          a,b,c
      -->
  ]]>
  foo, foo
   <![CDATA[
      <!--
          d,e,f
      -->
  ]]>
</style>

The content of the <style> element is this string:

      blah, blah
      <!--
          a,b,c
      -->
      foo, foo
      <!--
          d,e,f
      -->

Notice that it appears there are two comments, but they are not, they are
"fake comments".

I want to get just the content inside the fake comments:

         a,b,c
         d,e,f

The following XPath expression seems to give the desired results:

for $i in tokenize(., '<!--') return if (contains($i, '-->')) then
substring-before($i, '-->') else ()

Is there a better XPath expression -- simpler, more precise, more complete?

/Roger