Subject: RE: Re: ANN: A static code quality tool, for XSLT code
From: "Michael Kay" <mike@xxxxxxxxxxxx>
Date: Thu, 1 Jan 2009 17:16:22 -0000
|
> Since you require XSLT 2.0 it would be more precise to use
> regular expression for stuff like this, like
>
> //@*[matches(., 'true[^(] | false[^(]', 'x')]
>
> (I like the x flag, which allows white-space for readability.)
yes, so long as you get the regexes right! This example only matches "true"
if it is followed by a character other than "(" - it fails to match
select="true" on its own.
To be honest, I think a production-quality tool that attempts this kind of
analysis should parse the XPath expressions properly.
Michael Kay
http://www.saxonica.com/
|