W3C home > Mailing lists > Public > www-ql@w3.org > July to September 2004

AW: Missing productions in Core BNF (WD 20 February 2004)?

From: Bog, Anja <anja.bog@sap.com>
Date: Thu, 26 Aug 2004 15:13:20 +0200
Message-ID: <61E587F9B3BEBB479E8684F54AD5A717872B3E@dewdfe23.wdf.sap.corp>
To: "'www-ql@w3.org'" <www-ql@w3.org>
Hi everybody,
 
there's another issue I am wondering about now: If you produce StepExpr from ValueExpr
 
>ValueExpr -> StepExpr -> PrimaryExpr 
 
how will you produce Sequences of AxisSteps like they are usual for PathExpressions?
 
Tks,
Anja

-----Ursprüngliche Nachricht-----
Von: www-ql-request@w3.org [mailto:www-ql-request@w3.org] Im Auftrag von Jerome Simeon
Gesendet: Mittwoch, 25. August 2004 20:10
An: Sievers, Jan
Cc: www-Ql (www-ql@w3.org)
Betreff: Re: Missing productions in Core BNF (WD 20 February 2004)?



Hi Jan, 

Yes, that is a bug. Good catch. The connection: 

ValueExpr -> PathExpr -> RelativePathExpr -> StepExpr -> FilterExpr -> PrimaryExpr 

was lost. It should be preserved in the XQuery core through: 

ValueExpr -> StepExpr -> PrimaryExpr 

There is no need for a FilterExpr which would just be a reference to PrimaryExpr, but 
we need to introduce back the StepExpr as being the union of AxisStep or PrimaryExpr. 

The corresponding grammar productions should be changed to the following: 

ValueExpr  ::=  ValidateExpr | StepExpr 
StepExpr  ::=  AxisStep | PrimaryExpr 

This will be fixed in the next version of the Formal Semantics document. 

Best, 
- Jerome 




"Sievers, Jan" <j.sievers@sap.com> 
Sent by: www-ql-request@w3.org 


08/25/2004 11:36 AM 


To
"www-Ql (www-ql@w3.org)" <www-ql@w3.org> 

cc

Subject
Missing productions in Core BNF (WD 20 February 2004)?

	





Hi everybody,

I am wondering if there are some productions missing in the Core BNF defined in "XQuery 1.0 and XPath 2.0 Formal Semantics" (Working Draft 20 February 2004).

For instance there is no "StepExpr" or "FilterExpr" and the production [53 (Core)] "PrimaryExpr ::= ..." is never used.

Is this a bug?
Or is there something special about the normalization rules?

Thanks,
Jan Sievers
Received on Thursday, 26 August 2004 13:14:04 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:17:16 UTC