NetNews Usenet Archive 1992 #26/NN_1992_26.iso/spool/comp/lang/c/16323

Internet Message Format | 1992-11-11 | 1.7 KB

Path: sparky!uunet!zaphod.mps.ohio-state.edu!uwm.edu!spool.mu.edu!agate!dog.ee.lbl.gov!horse.ee.lbl.gov!torek
From: [e-mail] (Chris Torek)
Newsgroups: comp.lang.c
Subject: Re: Hows this for odd
Date: 11 Nov 1992 12:23:47 GMT
Organization: Lawrence Berkeley Laboratory, Berkeley
Lines: 31
Message-ID: <[e-mail]>
References: <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Reply-To: [e-mail] (Chris Torek)
NNTP-Posting-Host: 128.3.112.15

(I was hoping to stay out of this mess, but...)

I *can* think of one good use for

    i == 0 ? TRUE : FALSE

(instead of just `i == 0', or, for the terse among us, `!i').  In this
one situation, however, I would probably spell `TRUE' and `FALSE'
differently.

The situation I have in mind is a C-language implementation of some
other (non-C) language, where the other language *has* a boolean type
with specific values for `true' and `false'.  For instance, a Pascal
system written in C might use something like:

    i == 0 ? V_TRUE : V_FALSE

to get the value nodes (which presumably would have some kind of
aggregate type, or be an index into a table) for the `true' and `false'
values.  These would be different nodes from, e.g., the integer
constant `0' and the untyped pointer `nil' (which might be V_0 and
V_NIL respectively).

The situation is not always clear-cut, and blanket statements like
`a definition for TRUE is always bad' are bound to be false sometimes.

(For the most part, though, I do think `i == 0 ? TRUE : FALSE' is at
best suspicious.)
--
In-Real-Life: Chris Torek, Lawrence Berkeley Lab CSE/EE ([phone])
Berkeley, CA        Domain:    [e-mail]