home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.std.c
- Path: sparky!uunet!cs.utexas.edu!torn!utzoo!telly!druid!pseudo!mjn
- From: mjn@pseudo.uucp (Murray Nesbitt)
- Subject: Re: is fflush(stdin) conformant???
- Organization: Private system in Toronto, ON
- Date: Thu, 5 Nov 1992 11:02:41 GMT
- Message-ID: <MJN.92Nov5030241@pseudo.uucp>
- In-Reply-To: mgk@sql.sybase.com's message of 2 Nov 92 21:18:55 GMT
- References: <MGK.92Nov2131855@cedar.sql.sybase.com>
- Sender: mjn@pseudo.uucp (Murray Nesbitt)
- Lines: 19
-
-
- mgk@sql.sybase.com (Michael Keirnan aka Joe Cool) writes:
-
- >I am not actually comtemplating using this construct. A coworker is
- >enrolled in a course where the instructor has advised flushing stdin
- >(yes, stdin) in conjunction with using scanf() to "flush the unwanted
- >newline." This seemed rather odd to me, but hey, what do I know?
- >So what sayeth the gospel on the subject?
-
- fflush()ing an input stream has no defined effect in ANSI C - it is
- intended for output streams or update streams (in which the most
- recent I/O operation was not input) only.
-
- It may "flush the unwanted newline" on some particular systems, but if
- the instructor claims it is a universal way of doing this in C, she's
- wrong.
-
- --
- Murray
-