Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!cmcl2!yale!mintaka!spdcc!ima!dirtydog!karl
From: karl@ima.isc.com (Karl Heuer)
Newsgroups: comp.lang.c
Subject: Re: sprintf badness
Keywords: sprintf, sscanf
Message-ID: <1991Jan21.212216.27724@dirtydog.ima.isc.com>
Date: 21 Jan 91 21:22:16 GMT
References: <473@bally.Bally.COM> <1991Jan21.004401.15257@nntp-server.caltech.edu>
Sender: news@dirtydog.ima.isc.com (NEWS ADMIN)
Distribution: usa
Organization: Interactive Systems
Lines: 19

In article <1991Jan21.004401.15257@nntp-server.caltech.edu> gceych@iago.caltech.edu writes:
>>  double k;
>>  sscanf(buf,"%4d%7f%4d",&i,&k,&j);
>                  ^^^
>I believe this should be "%7lf", for a DOUBLE number.

Right.  This should fix the problem.

>Same mistake is made in the next line of the program.
>>  printf("i = %d  j = %d  k= %7.2f",i,j,k);

Wrong.  For printf(), "%f" is correct and "%lf" is illegal.  This unfortunate
asymmetry is a result of history and the default widening rules.

(You'll probably find that "%lf" works on your local implementation of
printf().  But since "%f" also works and is Standard, that's the one to use.
Fix it now, before you port to an implementation that works differently.)

Karl W. Z. Heuer (karl@ima.isc.com or uunet!ima!karl), The Walking Lint