Date: Thu, 7 Mar 2002 02:16:30 -0500 From: Brian T.Schellenberger <bts@babbleon.org> To: Poul-Henning Kamp <phk@critter.freebsd.dk>, obrien@FreeBSD.ORG Cc: Garance A Drosihn <drosih@rpi.edu>, Mike Meyer <mwm-dated-1015843484.1eabc5@mired.org>, hackers@FreeBSD.ORG Subject: Re: RFC: style(9) isn't explicit about booleans for testing -- an actual analysis of the code! Message-ID: <20020307071630.E26DBBA03@i8k.babbleon.org> In-Reply-To: <52204.1015480748@critter.freebsd.dk> References: <52204.1015480748@critter.freebsd.dk>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thursday 07 March 2002 12:59 am, Poul-Henning Kamp wrote:
| In message <20020306192234.B55297@dragon.nuxi.com>, "David O'Brien" writes:
| >Implies??? I thought I was quite explicit:
| >
| > to prevent is "if (!strcmp(a,b))" which when read is extremely wrong
| > of that is actually happening.
| >
| >! is pronounced "NOT". When read "if not string compare a with b then do
| > X", is the opposite of the the logic of the expression does. Which is
| > "if string compare a with b is equal then do X". ["if (strcmp(a,b) ==
| > 0)"]
|
| Well, we're clearly into "IMO" land here, so lets ignore that :-)
No, we aren't, and let's not.
Maybe your brain has gotten used to it, but to us ordinary mortals, even us
ordinary mortals who've been slogging C code for time periods that can be
measured in decades (yikes!), it is very tempting to read
i f (!strcmp(a,b,l))
as "if the strings don't compare" which , in ordinary usage, means don't
compare *equal*. But of course the C program is going to proceed to do
exactly the opposite of that.
Now, I personally find the constructions
if (!p) do_thing_for_null_pointer;
and
if (p) do_thing_for_valid_pointer;
quite readable and would prefer that an abstract Guide To Perfect Style
blessed them since I find them quite readable as "if I don't have a pointer"
and "if I do have a pointer".
But *ALL* of this is beside the nominal point ANYWAY, which is to discuss
the proper wording for the man(9) style guide which is supposed to document
how things things are actually done in the kernel, not personal preference.
===================================================================
So I just grepped the code
(find /usr/src -type f -exec grep 'if.*strcmp' {} \;)
and analyize the results.
Overall, there are 1831 !strcmp test and 3363 strcmp.*==0 tests
Within the kernel there are 84 !strcmp tests and 155 strcmp.*== 0 tests
So 65% either way in favor of the proposed change in wording -- a 2:1
majority of the code votes FOR the proposed change.
(This is subject to minor error given that it's all based on greps and
statistics, but I did the analysis a few different ways using different
plausible regular expressions to determine the two code paths, and all of
analysis showed that it was at least 2:1 using the == / != 0 tests.)
--
Brian T. Schellenberger . . . . . . . bts@wnt.sas.com (work)
Brian, the man from Babble-On . . . . bts@babbleon.org (personal)
ME --> http://www.babbleon.org
http://www.eff.org <-- GOOD GUYS --> http://www.programming-freedom.org
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020307071630.E26DBBA03>