Re: Top 10 most common hard skills listed on resumes...

Tim Rentsch <tr.17687@z991.linuxsc.com> writes:

Keith Thompson <Keith.S.Thompson+u@gmail.com> writes:

Bart <bc@freeuk.com> writes:

[...]

I can also say that the C grammar is buggy:
>
    assignment-expression:
        conditional-expression
        unary-expression asssignment-operator assignment-expression
>
When attempting to parse an assignment-expression, do you go for a
conditional-expression or unary-expression?
>
The latter is a subset of the former.  If you go for a
conditional-expression and find that an assignment-operator
follows, now you have to perform some analysis on the LHS to see
if that conditional-expression contains only a unary-expression.
[...]

>
[...] I'm skeptical that the C grammar is buggy.  [...]

>
It appears that what Bart means by buggy is different from what
you mean.  I think what Bart means is that the grammar is not
suitable for being used by a particular parsing algorithm.  Of
course that is not what the C standard means to supply, which is
rules of grammar that exactly reflect what syntactic forms are
suitable (syntactically) as C programs, without regard to how
input source is processed.  The C standard's rules of grammar
are meant as a declarative specification, not as a procedural
description.  Bart's complaint is, I believe, a complaint that
the grammar rules in the C standard do not provide a suitable
procedural description (for the procedural framework he has in
mind).  They aren't meant to.  Your comment is meant, I believe,
to be about whether the grammar rules in the C standard provide
an accurate declarative specification, which they almost
certainly do, especially as regards to the limited area of
expression syntax.

One data point: the yacc/bison grammar I cited upthread at
<https://gist.github.com/codebrainz/2933703>, which as far as I can
tell *is* meant as a procedural description, includes the following:

assignment_expression
    : conditional_expression
    | unary_expression assignment_operator assignment_expression
    ;

I haven't tested it beyond running `bison c99.y`, nor have I compared
the rest of c99.y (and c99.l) to the grammar in the standard.

I'm (still) skeptical that Bart has found a bug in the ISO C grammar
in any reasonable sense of the word "bug".

My impression at this point is that the ISO C grammar can be used
as-is to construct a working parser by a straightforward translation.
There are good reasons for not doing so (producing user-friendlier
error messages).  This is only an impression, and I could easily
be wrong.

I'll also note that you've discussed what you think Bart meant,
but not whether he's correct.

--
Keith Thompson (The_Other_Keith) Keith.S.Thompson+u@gmail.com
void Void(void) { Void(); } /* The recursive call of the void */