- From: Michael Dyck <jmdyck@ibiblio.org>
- Date: Fri, 11 Nov 2016 09:43:56 -0500
- To: public-xsl-query@w3.org
> Decision: Adopt proposed resolution for Bug 29959 > Action A-660-08: mdyck to incorporate resolution of Bug 29959 in the grammar. First, this is a typo in the minutes: both occurrences of "29959" here should be "29989". Second, "if a decision is recorded about the resolution of a bug, the decision must state exactly what the resolution is", and this decision does not. (Nor is it possible possible to infer the resolution from the discussion in the minutes, as there's none recorded.) So I will state for the record: For Bug 29989 ([XQ31] CurlyArrayConstructor does not use EnclosedExpr), the agreed resolution was to make the change suggested in comment 0 of the bug. (Also, although the bug was raised against XQuery 3.1, the same production occurs in XPath 3.1, and it seems reasonable to make the change in both grammars.) ---- So, in the xpath31 and xquery31 grammars, I have changed: CurlyArrayConstructor ::= "array" "{" Expr? "}" to: CurlyArrayConstructor ::= "array" EnclosedExpr I have also made the necessary changes in the converter-to-XQueryX, and regenerated the applets, and committed all that. ---- Note that the documents in: xquery-31/ xpath-full-text-31/ xquery-update-31/ will need to be rebuilt against the new grammar. In the XPath/XQuery 3.1 docs, the prose around CurlyArrayConstructor is loose enough that it probably doesn't need to be changed, but Jonathan should maybe look at it anyway. As Josh noted in the bug, the semantics of EnclosedExpr now define the semantics of "array {}", so that needn't be specified separately. But an example to that effect would probably be a good idea. -Michael
Received on Friday, 11 November 2016 14:44:35 UTC