NetNews Usenet Archive 1993 #3/NN_1993_3.iso/spool/comp/lang/c/20266

Internet Message Format | 1993-01-28 | 1.6 KB

Xref: sparky comp.lang.c:20266 comp.std.c:3430
Newsgroups: comp.lang.c,comp.std.c
Path: sparky!uunet!sangam!cmie!rs
From: [e-mail] (Rajappa Iyer)
Subject: Re: Preprocessor question
Message-ID: <[e-mail]>
Keywords: Preprocessor, string
Organization: Centre for Monitoring Indian Economy, Bombay
References: <[e-mail]>
Date: Wed, 27 Jan 1993 07:28:14 GMT
Lines: 35

[e-mail] (Karim Younes) writes:

>I am trying to have a #defined variable recognized by the
>preprocessor inside a string. In other words, I want to
>do something like this:

>#define JUNK 50
>static char *junk = "JUNK ways to leave your lover."

The only ANSI way I can think of doing it is (yecch!):

#define JUNK "50"
static char *junk = JUNK " ways to leave your lover."

Interestingly, I tried the following with gcc 2.2.2:

#define JUNK    50
#define STR(x)    #x

char *junk = STR(JUNK) " ways to leave your lover."

junk expands to "JUNK ways to leave your lover." Which kind of
makes me question the utility of the `#' preprocessor operator.
K&R 2 says that "... replacement token sequence is preceded by #,
string quotes (") are placed around the corresponding parameter
and then both the # and the parameter identifier are replaced by
the quoted argument."

In other words, in the above example `x' is replaced by `JUNK'
and quoted---`JUNK' is not expanded. Is this indeed the correct
behavior? If it is, I really think it severely restricts the
utility of the `#' operator. Any comments, anyone?
--
Rajappa Iyer            [e-mail] (Preferred)
                [e-mail]