Subject: Re: [xsl] Saxon Quirk From: "Dave Hartnoll" <themainman@xxxxxxx> Date: Thu, 15 Feb 2001 17:05:42 -0000 |
> > Instant Saxon 6.0.2/6.1/6.2 and Xalan 1.2.2. seem to be on better terms > > with preceding-sibling:: then with preceding:: > > > > <xsl:variable name='allpubs' > > select="//pubtag[@list='yes']/@pub[not(. = > > ../preceding-sibling::pubtag/@pub)]" /> > This is true for the simple example that I provided, but my Real World > project has the "pubtag" at any level, so preceding-sibling would not work > for me. I've done some tests and it seems that the preceding:: axis is broken in Saxon if the context node is an attribute node. You can get round it by doing: ../preceding::pubtag/ I can't figure out why preceding-sibling works, though. Perhaps there's something wrong with that too? I'll submit a clearer test to the Saxon list which also shows the fault with preceding. Dave Hartnoll. XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
RE: [xsl] Saxon Quirk, Devlin, Kurt | Thread | RE: [xsl] Saxon Quirk, Michael Kay |
Re: [xsl] XSLT 1.1 comments - the c, Francis Norton | Date | Re: [xsl] [ANNOUNCE] XSLT-process 1, Sebastian Rahtz |
Month |