Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!mips!kpc.com!ardent!jss
From: jss@gold.kpc.com (Jerry Schwarz)
Newsgroups: comp.lang.c++
Subject: Re: istream::putback(char c)
Message-ID: <JSS.91Apr9182940@gold.kpc.com>
Date: 10 Apr 91 01:29:40 GMT
References: <71302@brunix.UUCP>
Sender: uucp@kpc.com (UNIX-to-UNIX Copy)
Organization: Kubota Pacific Computer
Lines: 38
In-Reply-To: sdm@cs.brown.edu's message of 8 Apr 91 19:00:07 GMT
Nntp-Posting-Host: gold

In article <71302@brunix.UUCP> sdm@cs.brown.edu (Scott Meyers) writes:

   From the istream man page:

	     insp=&ins.putback(c)
		  Attempts to back up ins.rdbuf().  c  must  be  the
		  character   before   ins.rdbuf()'s   get  pointer.
		  (Unless other activity  is  modifying  ins.rdbuf()
		  this is the last character extracted from ins.) If
		  it is not, the effect is undefined.   

This peculiar interface was motivated by a desire to make it easy to
derive new streambufs.

>       1.  Why am I limited to putting back the last character I removed?  
>	   There is no such limitation for ungetc in the standard C library.

It is possible that you are extracting characters from a readonly data
structure.  For example a strstreambuf that was initialized with a
|const char*|.  strstreambuf ignores the argument and just backs up
the (internal) pointer.


>       2.  Given that I am limited to putting back the last character I 
>	   read, why does putback take a parameter?

Putback was originally specified in a way that corresponded to ungetc.
Some streambufs (e.g. filebuf) will use the character and do allow you
to putback a character other than the one you extracted most recently.
The restriction (that you should only putback the previously extracted
character) was added later when I noticed the situation described in
case 1.

The X3J16 library working group is considering eliminating the
argument.  That is declaraing istream::putback() to back up the
pointer.

Jerry Schwarz