Comments on: XPath puzzler http://dubinko.info/blog/2007/12/15/xpath-puzzler/ From an XML geek, a reader, a writer, a connector, a man of the people (says keep hope alive) Sat, 01 Oct 2011 16:18:55 +0000 hourly 1 http://wordpress.org/?v=3.3.2 By: Push Button Paradise | Blog Archive | XPath puzzler: solution http://dubinko.info/blog/2007/12/15/xpath-puzzler/comment-page-1/#comment-3843 Push Button Paradise | Blog Archive | XPath puzzler: solution Mon, 31 Dec 2007 07:39:50 +0000 http://dubinko.info/blog/2007/12/15/xpath-puzzler/#comment-3843 [...] to all the folks who showed interest in this little XPath puzzler published here a few weeks ago. Some asked to see the dataset, but I’m not able to release it [...] [...] to all the folks who showed interest in this little XPath puzzler published here a few weeks ago. Some asked to see the dataset, but I’m not able to release it [...]

]]> By: Anthony B. Coates http://dubinko.info/blog/2007/12/15/xpath-puzzler/comment-page-1/#comment-3829 Anthony B. Coates Wed, 19 Dec 2007 17:08:35 +0000 http://dubinko.info/blog/2007/12/15/xpath-puzzler/#comment-3829 Can you post the dataset (with the data text replaced by some bogus text if necessary to protect the contents)? Cheers, Tony. Can you post the dataset (with the data text replaced by some bogus text if necessary to protect the contents)?
Cheers, Tony.

]]> By: Tom Passin http://dubinko.info/blog/2007/12/15/xpath-puzzler/comment-page-1/#comment-3827 Tom Passin Sun, 16 Dec 2007 23:48:02 +0000 http://dubinko.info/blog/2007/12/15/xpath-puzzler/#comment-3827 You would get this if some of the meta elements had more that one property attribute. Since that wouldn't be well-formed, and the xslt processor shouldn't be receiving multiple attributes of the same name, let alone be counting them, it would indicate one or more bugs, in the parser and in the xslt processor. You would get this if some of the meta elements had more that one property attribute. Since that wouldn’t be well-formed, and the xslt processor shouldn’t be receiving multiple attributes of the same name, let alone be counting them, it would indicate one or more bugs, in the parser and in the xslt processor.

]]> By: Dimitre Novatchev http://dubinko.info/blog/2007/12/15/xpath-puzzler/comment-page-1/#comment-3826 Dimitre Novatchev Sun, 16 Dec 2007 15:53:09 +0000 http://dubinko.info/blog/2007/12/15/xpath-puzzler/#comment-3826 Do try with different XPath engines in order to determine if the observed result is only produced by your XPath engine, in which case this is a clear indication of a bug. If most/all XPath engines report the observed result, could you, please, provide the xml file against which the XPath expression is evaluated? Cheers, Dimitre Novatchev Do try with different XPath engines in order to determine if the observed result is only produced by your XPath engine, in which case this is a clear indication of a bug. If most/all XPath engines report the observed result, could you, please, provide the xml file against which the XPath expression is evaluated?

Cheers,
Dimitre Novatchev

]]> By: M. David Peterson http://dubinko.info/blog/2007/12/15/xpath-puzzler/comment-page-1/#comment-3825 M. David Peterson Sun, 16 Dec 2007 07:26:47 +0000 http://dubinko.info/blog/2007/12/15/xpath-puzzler/#comment-3825 Are the property attributes potentially in different namespaces, and your XPath engine is returning all matches on local-name() as opposed to name()? It's possible that <meta foo:property="bar" bar:property="baz"/> on one or two of your meta elements is causing the increase. If you try //meta[@property] your count should be something equal to or less than the total count of the meta elements. If it's not, then there's really something weird going on. Are the property attributes potentially in different namespaces, and your XPath engine is returning all matches on local-name() as opposed to name()? It’s possible that <meta foo:property=”bar” bar:property=”baz”/> on one or two of your meta elements is causing the increase.

If you try //meta[@property] your count should be something equal to or less than the total count of the meta elements. If it’s not, then there’s really something weird going on.

]]>