- From: Jonathan Robie <jonathan.robie@datadirect.com>
- Date: Mon, 08 Nov 2004 07:21:20 -0500
- To: Arjan Huijzer <huijzer@gmail.com>
- CC: www-ql@w3.org
Hi Arjan, The change log in I.2 is pretty informal in the way it discusses this change. Take a look at 3.2 to see the actual change. Look for the text starting with "Each operation E1/E2 is evaluated as follows:". The change is that it is no longer an error if E2 evaluates to a sequence of atomic values. We provide specifically for that case: > If every evaluation of E2 returns a (possibly empty) sequence of > atomic values, these sequences are concatenated, in order, and > returned. Although the change log is informal, I think this reference is precise and clear. Do you agree? Jonathan Arjan Huijzer wrote: > Hello, > > I have a question about the usage of XPath expressions. > > In the latest XQuery specifications it says that the last step of a > XPath expression can now return a sequence of nodes (I.2 29 Oct > 2004). Looking at the grammar rules (A.1 EBNF), it seems that > PathExpr (rule 61) is a special form of an OrExpr (43), which can > only return a "true" or "false" value. > > Are the specifications not clear on this point or am I missing > something? > > Thanks in advance for any help! > > Arjan Huijzer The Netherlands >
Received on Monday, 8 November 2004 12:21:59 UTC