Re: XQuery lexical stuff

Michael Dyck <michaeldyck@shaw.ca> wrote:
> One advantage of the CFG
> over the PDA (in my opinion) is that it's easier to spot the mistakes
> in the CFG.

I wonder if you could define the PDA/State Transitions in terms of a CFG,
for the purposes of design, and perhaps also to use in the documentation?
This would be worth exploring.

The reason the PDA is valuable is because it works with tools like JavaCC
and Lex.  It is very important to us that this grammar be easily expressed
by these tools.

> Ultimately, I wanted to question the need for the XQuery spec to define
> a tokenizer at all.

I brought this up at the last F2F in Cannes.  People were interested in the
idea of doing less tokenization-level specification.  We discussed making
most of the lexical section into a non-normative appendix.  In other words,
this is how we implemented and verified it, but implementors mayk obviously
do what they want.

You sound like you know more than the average guy about parsing.  Do you
have any special credentials in this area?  The reason I ask is we're
looking for "parser academians"  to look at the grammar (XPath and XQuery)
and brainstorm with us on ways to make it better.  Of special issue are
so-called long tokens, lexical states, and the issue of whether or not
there should be reserved words in the grammar.

-scott

Received on Thursday, 21 March 2002 06:13:26 UTC