[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: Ambiguity in XML spec

• From: "David Brownell" <david-b@p...>
• To: "Brown, Bryan" <bryanb@u...>, <xml-dev@x...>
• Date: Thu, 18 May 2000 08:58:33 +0200

The BNF needs more than one token of lookahead ... it's
not LALR(1), LL(1), etc.

The fix is to morph the grammar productions till
you get a rewritten grammar that does have the sort
of characteristics that your parser requires.  (But
one that still recognizes the same language!!)

Most compiler texts will talk about such issues; it's
been common that the grammars are written to be readily
understood by people, but computer tools often need
just a bit more help to get their jobs done.

- Dave


----- Original Message -----
From: Brown, Bryan <bryanb@u...>
To: <xml-dev@x...>
Sent: Thursday, May 18, 2000 12:15 AM
Subject: Ambiguity in XML spec


> I am currently confused on part of the XML spec. It appears to me that
> production 82 (NotationDecl) is ambiguous. I can not find any mention in
the
> XML errata, and the editor of the spec (Tim bray) hasn't responded.
>
> Here is my thinking
>
> if you are parsing a NotationDecl after you read the name and
> optional whitespace you need to determine if the stuff next is an
ExternalID
> (production 75) or a PublicID (production 83).
> So you check the next token which is 'PUBLIC',
>
> unfortunately both the ExternalID and PublicID may begin with
> 'PUBLIC', further if it is an ExternalID after 'PUBLIC' would follow S
> PubidLiteral S SystemLiteral but after a PublicID follows S PubidLiteral.
>
> So after reading 'PUBLIC' you have no idea if you are looking for "S
> S PubidLiteral S SystemLiteral " or just "S PubidLiteral"
>
> What am i missing ?
>
> Bryan
>
>
***************************************************************************
> This is xml-dev, the mailing list for XML developers.
> To unsubscribe, mailto:majordomo@x...&BODY=unsubscribe%20xml-dev
> List archives are available at http://xml.org/archives/xml-dev/
>
***************************************************************************


***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@x...&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/
***************************************************************************

• References:
  • Ambiguity in XML spec
    • From: "Brown, Bryan" <bryanb@u...>

• Prev by Date: land specific and international use of common labels
• Next by Date: Re: Irony heaped on irony
• Previous by thread: Ambiguity in XML spec
• Next by thread: Re: Ambiguity in XML spec
• Index(es):
  • Date
  • Thread