[Monetdb-developers] [Monetdb-pf-checkins] pathfinder/compiler/debug abssynprint.c, 1.25, 1.26 coreprint.c, 1.26, 1.27

Ying Zhang Y.Zhang at cwi.nl
Fri Oct 27 12:45:16 CEST 2006


On Fri, Oct 27, 2006 at 11:26:48AM +0200, Torsten Grust wrote:
> Hi all,
> 
> 
> On Oct 27, 2006, 11:12 AM, Ying Zhang wrote with possible deletions:
> >[...]
> >Would you please change the implementation of the syntax according to
> >the discussion on the syntax?  The final version (I hope:)) of the
> >syntax should be:
> >
> >    "execute at" Expr "xquery" FunctionCall
> 
> I have not seen the discussion (why? -- has it been conducted on  
> monetdb-developers), 

Obviously forgotten.  I have just CC-ed the e-mail to the developers
list.

> but I strongly vote to leave the syntax as is,  
> i.e., use
> 
> 	execute at { Expr } { FunctionCall }
> 
> This is in line with other similar syntactic constructs in XQuery,  
> foremost the computed document, attribute, element, text, comment  
> constructors, the ordered/unordered specifiers, and the validate  
> expression.  Much like unordered { e } is a hint to the system to  
> evaluate expression with lax ordering requirements, execute at  
> { foo } { e } is an instruction to the system to evaluate e at foo.
> 
> Much like for the element/attribute construtors, I also propose to  
> admit the syntactically lighter
> 
> 	execute at URILiteral { Expr }
                            ^^^^

But we only want to support FunctionCall here, instead of an arbitrary
xquery expression...


Jennie

> 
> as an additional syntactic variant whenever the remote system URI  
> need not be dynamically computed.
> 
> Cheers,
>    --Teggy
> -- 
> | Torsten Grust
> | torsten.grust at gmail.com
> 
> 
> 




More information about the developers-list mailing list