Re: Appendix F regular expression (XmlCharRef)

Then I don't understand the reason for rule [19]
in Appendix F, since the character "A" can always
be represented by using &-#-6-5-;  or &-#-x-4-1-;
The stiring is converted to the character it
represents by the xml parser before the schema
routines ever see it.  No special rule is required.


"Biron,Paul V" wrote:

>
>
> > -----Original Message-----
> > From: tmoog [mailto:tmoog@sarvega.com]
> > Sent: Mon, Nov 17, 2003 06:05
> > To: xmlschema-dev@w3.org
> >
> > a.  Schema Part 2, Appendix F, production [10] is a
> > definition of "normal character".  This appears to disagree
> > with the definition of "normal character"
> > and "meta-character" in the text immediately preceding in
> > that it omits left and right curly braces.  I don't see any
> > correction of this in the errata.  Have I missed something ?
>
> this is a known problem and I thought there was already an erratum on
> this...but I see that the proposed errata text has never been approved
> by the WG...we'll get to it soon.
>
> > b.  Suppose a malformed XmlCharRef appears in a charRange
> > (e.g. &-#-1-2-x-semicolon instead of &-#-1-2-3-semicolon).
> > Since neither "&" (nor "#" nor ";") are meta-characters, this
> > seems to imply that the sequence should be interpreted as six
> > ordinary XmlChar instead of causing an error ("Malformed
> > XmlCharRef in regular expression at line ...").
> > Is this interpretation correct ?
>
> No, that interp is not correct.  The reason is that:
>
>         <x>&#12x;</x>
>
> is not well-formed.
>
> pvb