Re: 2E PER: Lexical Representation of unsignedLong, unsignedInt, etc...

At 5:50 PM -0400 10/20/04, Michael Glavassevich wrote:

>Is it really true that an optional sign is allowed in the lexical form of
>the unsigned integer types?  If so shouldn't that be stated in the
>description of the lexical representation for each of those types?  If not
>shouldn't unsignedLong specify a pattern facet which disallows a leading
>sign?  There appears to be an error either in the schema or in the lexical
>representation descriptions.

Quite right:  The description doesn't match the specified derivation.
I can't speak for the WG--I suspect there will be a quick survey of
implementations to see if they commonly accept an optional leading
'+'.  However, since unsigned integer allows the sign(s), if I had
to I would guess that the others will too.

In any case, I'm sure the WG will make a decision one way or the other.
-- 
Dave Peterson
SGMLWorks!

davep@iit.edu

Received on Wednesday, 20 October 2004 23:41:22 UTC