Re: what I learned from today's discussion of delimiters

Liam R. E. Quin writes:

> On Tue, 2022-01-25 at 21:26 -0700, C. M. Sperberg-McQueen wrote:
>> 
>> (3) Neither using braces in pragma delimiters nor avoiding them will
>>     have any major technical effect on the spec.
>
> If you want to put them into element content in XSLT 3 or XQuery, curly
> braces are of course special. You can get around it in XSLT with
> expand-text="no" and in XQuery with literal element constructors,
> ``[ stuff here ]``
> as long as "stuff here" doesn't contain `{ ... }`.

All true.

I think, however. that if different choices affect the convenience of
people who want to write ixml grammars in XQuery or XSLT, that effect
will be external to the ixml spec; it will not mean the ixml spec is
more or less complex than it would be had we made a different choice.

I find remembering to duplicate braces when writing ixml inside an
XQuery module a bit tedious, but not an insuperable difficulty.  When I
want to do serious work in ixml using XQuery or XSLT, i use the XML form
of grammars, not the ixml form.

And of course the choice of delimiters for pragmas won't be decisive for
the ease or difficulty of writing ixml in XQuery or XSLT contexts,
because braces are already prominent in ixml syntax; even if we choose
non-brace delimiters for pragmas, braces will remain as the delimiters
for comments.  If the special role of braces in XQuery and XSLT makes
them a bad choice for important syntactic roles in ixml, then that
battle has already been lost.

> XQuery uses (#...#) with an EQname at the start,
>   "(#" S? stuff  "#)"
> where stuff is a sequence of Char not including "#)"

> It might be nice to be the same if that's possible, instead of
> inventing something new.

You may be right, though I suspect that (# ... #) would please no one:
it doesn't use braces so it will not appeal to anyone who would like
pragmas to resemble comments visually, and it doesn't use
single-character delimiters so it will not appeal to anyone who would
like very short pragmas to be feasible.  It might be adopted if everyone
decides that we would rather have a situation in which nobody 'wins' the
argument and everyone is unhappy about the result, though for different
reasons.

Michael


-- 
C. M. Sperberg-McQueen
Black Mesa Technologies LLC
http://blackmesatech.com

Received on Wednesday, 26 January 2022 05:29:08 UTC