Re: <shape> datatype for clip

Le Saturday 19 October 2002 01:59, Oleg Tkachenko a écrit:
> Éric Bischoff wrote:
> > Another one that keeps bugging me :
> >
> > -------------------------------------------------------------------------
> > The BNC production for a function call (section 5.9.4) is :
> >
> > 	[3]   FunctionCall   ::=
> > 	FunctionName '(' ( Argument ( ',' Argument)*)? ')'
> >
> > as you can see, it does not mention any whitespace, neither directly nor
> > indirectly through FunctionName and Argument productions.
> >
> > At the same time, in section 5.9.11, it is said that
> >
> > 	If the character following an NCName (possibly after intervening
> > 	ExprWhitespace) is "(", then the token must be recognized as
> > FunctionName.
> >
> > which explicitly says one could find white space. This is contradictory.
> > -------------------------------------------------------------------------
>
> It reminds me xpath lexical structure definition, which explicitly
> states that whitespace can be freely used between tokens, and xsl spec
> actually defines the same: (...)

I also assumed in my implementation that whitespace is allowed here. If it is 
really allowed, as I and Oleg both seem to think, then the explanations from 
section 5.9.4 should reflect it:

[3]   FunctionCall   ::=
 	FunctionName S '(' S ( Argument S ( ',' S Argument)* S )? ')'

or something alike (I probably did not give it enough thought).

I have noticed a bunch of similar inconsistencies while reading the 
specification. Too bad I did not make a list, and now I don't remember them 
anymore :-(. Do you want me to forward them to xsl-editors@w3.org when I 
happen to notice them or remember them again?

-- 
- Linux produces remarkedly less hot air than Windows: under
Windows, the processor gets hot after just a few minutes...
- Yes, but it never stays on long enough to burn out!