Subject: Re: Pattern question: first child of first descendant
From: "Christian Roth" <roth@xxxxxxxxxxxxxx>
Date: Tue, 24 Mar 2009 02:52:00 +0100
|
>But predicates aren't, so how about this:
>
>entry[ancestor::table][1]/*[self::title or self::para][1]
Unfortunately not - it has the same issue as Syd's initial pattern. The
first part does a "match all entry elements that have a table ancestor
and that are the first child of some element", which again matches all
first entry children of all the row elements in the table.
-Christian
|