[XSL-LIST Mailing List Archive Home]
[By Thread]
[By Date]
On 15/02/2011 11:43, Andriy Gerasika wrote:
For this refactoring the key information needed to make it safe is not the type of the expression bound to the variable, but the way in which the variable is used, which can be analyzed using Saxon's expression tree - but it's by no means easy.
Re: [xsl] XSLT Lint 0.1
Subject: Re: [xsl] XSLT Lint 0.1 From: Michael Kay <mike@xxxxxxxxxxxx> Date: Tue, 15 Feb 2011 11:51:39 +0000 |
On 15/02/2011 11:43, Andriy Gerasika wrote:
On 02/14/2011 06:20 PM, Michael Kay wrote:But I think FindBugs does some quite deep semantic analysis before making its suggestions.
I will split "xsl:variable/xsl:value-of" rule in two:
"xsl:variable/xsl:value-of (safe)" will use Saxon API to check type of XPath expression
"xsl:variable/xsl:value-of (not safe)" will refactor all cases of xsl:variable/xsl:value-of, assuming human will choose correct refactorings in file/compare tool
For this refactoring the key information needed to make it safe is not the type of the expression bound to the variable, but the way in which the variable is used, which can be analyzed using Saxon's expression tree - but it's by no means easy.
Michael Kay Saxonica
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
Re: [xsl] XSLT Lint 0.1, Andriy Gerasika | Thread | [xsl] Unescaping URIs?, Eliot Kimber |
Re: [xsl] XSLT Lint 0.1, Andriy Gerasika | Date | [xsl] Saxon -dtd:off, Vasu Chakkera |
Month |
Keywords