xml-dev - Re: Ambiguity in XML spec

[richard@cogsci.ed.ac.uk (Richard Tobin)]

In article <eshaisg6indaqmicvoif0dtiunmjcoagd4@4ax.com>,
Steve Schafer <pandeng@telepath.com> wrote:

>>The grammar in the spec describes the language correctly, and the
>>productions in question can be used directly with compiler-generators
>>like yacc.
>
>That's simply not true. yacc generates parsers for LALR(1) grammars,
>and the grammar in the XML spec is not LALR(1), unless, as I mentioned
>earlier, you omit at least some of the regular expression-based
>productions.

That's why I said "the productions in question".  The problem of
context-dependent tokenisation is a more general one, affecting
many productions.  All the DTD keywords have this problem.

Of course, with yacc and lex you can use "lexical tie-ins" to handle
the context-dependent tokenisation.

-- Richard
-- 
Spam filter: to mail me from a .com/.net site, put my surname in the headers.

"The Internet is really just a series of bottlenecks joined by high
speed networks." - Sam Wilson

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