- From: <bugzilla@jessica.w3.org>
- Date: Wed, 16 Jul 2014 13:09:38 +0000
- To: public-qt-comments@w3.org
https://www.w3.org/Bugs/Public/show_bug.cgi?id=25316 --- Comment #6 from Abel Braaksma <abel.braaksma@xs4all.nl> --- 3) ... <about fn:position() > > Not sure of your point? As it was originally written, both rules under 19.8.7.8 resulted in crawling. Now that the first ends up in striding, the issue is moot, because there is a difference now. 5) ... <about fn:last() > > Not sure of your point... In the rule, you say (my words) "except if it contains focus-dependent functions... except last". The double exception is redundant, you disallow focus-dependent functions, fn:last() is prohibited anyway (by its own rules), so why making an exception to the exception (allowing it), only to find out during analysis that using fn:last will yield free-ranging anyway? Or does an expression exist that uses last() on a streaming node-set that is crawling and is allowed? On (4), the focus-setting container, I see now that the "container" is the whole construct, under the definition in the spec we write: " if an expression appears within the predicate of a filter expression, its focus-setting container is the filter expression.". I was confused with what we define is the "controlling operand". A suggestion -- I think this says the same, and it might be clearer: * Neither P, nor any operand of P at any depth that has B as its [controlling operand], is a context item expression, an axis expression, or a call on a focus-dependent function. [...] 7) I just noticed the "or" at the end of the second bullet point, I think this should be "and". -- You are receiving this mail because: You are the QA Contact for the bug.
Received on Wednesday, 16 July 2014 13:09:40 UTC