bison-patches
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Dubious features


From: Akim Demaille
Subject: Re: Dubious features
Date: Sat, 10 Jun 2006 16:34:20 +0200


Le 9 juin 06 à 00:13, Paul Eggert a écrit :

Akim Demaille <address@hidden> writes:

Are there any other Yaccs presenting this feature?

Sure.  I just checked OpenBSD 3.4, and its yacc has this extension.
And its yacc did the right thing with a comment containing "{", too,
so Bison is compatible with its yacc.

I'm not too surprised: if I'm correct there yacc and bison used to
be the same program, and that was implemented in the original grammar
parser that was written by hand.  I don't think that counts.  I have
tried to find some documentation to see if this is documented, but
I found nothing conclusive in either direction.

The proposed patch would break
this compatibility.

In theory it would, I'm not sure it does for real in the sense that
I don't know grammars that use this feature.  And I very much doubt
some exist, but with a commented }.

This feature was removed from several versions of Bison, restored
by you on Christmas Eve 2002.  In the meantime, no one complained.


There has been many Bison which did not have it, and no one ever
complained.

Often people don't complain.  They just use yacc instead, and don't
bother with Bison.  See
<http://lists.gnu.org/archive/html/bison-patches/2006-06/ msg00031.html>
for the most recent example of this, dated yesterday.

Yes, I've seen that message, but given that nothing is said about the
problem, I'm not sure what we can conclude from that.


But this doesn't mean we should stop worrying about yacc
compatibility; quite the contrary.

I care about compatibility, but I'm more interest in gaining new
users thanks to new features rather that keeping users of old
stuff.

And in the present case I'm not asking for the removal of the
feature, but to make it what it should be: either rational or
context free.  As it is implemented the scanner is implementing
sort of a parser just for this case, that not elegant.  My
patch makes it completely rational.

But again, I'm curious to see uses of this.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]