Let me think this one over. I leave features undocumented sometimes for
various reasons: to keep my options open, because I have not had time to
test them, etc.
If it's in metag.bnf, that suggest that I think of it as supported. For
two reasons: Frist, I consider metag.bnf as a semi-public example. And
second, if it goes away, because of the circular dependency, I am in big
trouble. -- jeffrey
On 01/10/2014 10:26 PM, Ruslan Shvedov wrote:
As in /sequence ::= item+ separator => *[,]*. /The doc says it's a
no-no, but the attached test passes. Which of them needs patching? :)
What happened was I tried /sequence ::= item+ separator => //','/ and
it didn't work, but then I saw /separator => *[,] */in metag.bnf and
thought it can be useful.
To summarise, unless it's undocumented on some special purpose, I
think allowing character classes and single-quoted strings as
separator values could be a good thing.
[1] The separator keyword is only valid for a quantified right side,
and its value must be a symbol of the grammar.
—
https://github.com/jeffreykegler/Marpa--R2/blob/master/cpan/pod/Scanless/DSL.pod
--
You received this message because you are subscribed to the Google
Groups "marpa parser" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.
--
You received this message because you are subscribed to the Google Groups "marpa
parser" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.
