Path: csiph.com!eternal-september.org!feeder.eternal-september.org!reader01.eternal-september.org!.POSTED!not-for-mail
From: Tim Rentsch <tr.17687@z991.linuxsc.com>
Newsgroups: comp.lang.c
Subject: Re: Comma operator in for() statements
Date: Fri, 27 Mar 2020 07:42:00 -0700
Organization: A noiseless patient Spider
Lines: 63
Message-ID: <86r1xd97wn.fsf@linuxsc.com>
References: <m9kd7f94epq2eumphiui7gf2ubc6p8ah8e@4ax.com> <1NIdG.90366$0hk.19720@fx08.am4> <0fc029c2-7036-4ca9-86af-15913f874a08@googlegroups.com> <20200324202535.68c64bb9be8f95153b947f4b@g{oogle}mail.com> <878sjp1gk1.fsf@bsb.me.uk> <20200325145934.5f39e56fc029b19da139cc0d@g{oogle}mail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Injection-Info: reader02.eternal-september.org; posting-host="02cbe0882c270f81c5b8b4aa13e82222"; logging-data="10934"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX187R5CPgOhGNTKp9U+zBVQWfX7eKgXjhdc="
User-Agent: Gnus/5.11 (Gnus v5.11) Emacs/22.4 (gnu/linux)
Cancel-Lock: sha1:qMjlrDPB96nDzVlH2jt9OSwf1YI= sha1:Nsazwohtmg3ujc4+g9tA3PGxPeg=
Xref: csiph.com comp.lang.c:150859

Anton Shepelev <anton.txt@g{oogle}mail.com> writes:

> Ben Bacarisse to Anton Shepelev:
>
>>> My negative criticism of C's `for' statement is that it
>>> is essentialy a rather complicated *functional*
>>> statement, because it comprises four pieces of code that
>>> are executed in a non-trivial order and with non-trivial
>>> relations.  Passing pieces of code to statements is
>>> functional programming.  In procedural programming, one
>>> passes data, that is, in our case, -- index boundaries.
>>
>> What about the body?
>
> Indeed.  I was wrong, and should have named three instead of
> four functional pieces of code.
>
>> What about the condition in an 'if' and the arms of an
>> 'if'.
>
> The codition of an `if' is not a piece of code passed to a
> statement because it is executed only once, and can be
> replaced by its pre-calculated truth value, taking into
> account short-circuit evaluation and C's vague separation
> between statements and expressions.
>
> The body (or bodies) of control statmenets are not, in my
> view, passed to statements at all.  Rather, the control-flow
> statements are mark-up around blocks of linear (non-control)
> code.  Operator braces {} are just mark up.
>
>> The condition and body of a while?
>
> Yes, `while' "takes" the loop invariant in the same sense as
> does `for', yet `while' is much simpler and much more
> intutive in that it "takes" only one piece of code, which
> can be kept strictly as control code (only checks, no
> assignments), and that it is executed in the intuitive order
> of appearance.  The same applies to do..while.
>
> One of the worst abuses of `while' is code for which many
> programmers praise C:
>
>   char *s, *d;
>   /* ... */
>   while (*d++ = *s++);
>
> It employs several means of obfuscation:
>
>   1.  loop-control statements with side effects,
>
>   2.  use of counter-intuitive post-increment operator on
>       both parts of an assibnment.
>
>   3.  confusion of logical condition and comparison with
>       zero.
>
>   4.  an empty body of the `while' statement.

It's idiomatic C.  Perhaps more importantly, C not only allows
such constructions but embraces them stylistically.  Such things
have been part of C since its earliest beginnings.  If someone
wants to write Pascal-style code they should write in Pascal.