Re: libstdc++/4083: input stream with field width() set reads too much to std::string

Re: libstdc++/4083: input stream with field width() set reads too much to std::string

[David Hedbor]

The following reply was made to PR libstdc++/4083; it has been noted by GNATS.

From: David Hedbor <david@caudium.net>
To: Benjamin Kosnik <bkoz@redhat.com>
Cc: gcc-gnats@gcc.gnu.org
Subject: Re: libstdc++/4083: input stream with field width() set reads too much to std::string
Date: 24 Aug 2001 12:21:21 -0700

 Benjamin Kosnik <bkoz@redhat.com> writes:
 
 > > I assume that in the sections above, 21.3.7.9 is concerning
 > > std::string and 27.6.1.2.3 is for char *?
 > 
 > yep
 > 
 > > Sorry for the incorrect bug report. 
 > 
 > no worries. I'll close this one.
 
 Ok, great. gcc 3.0 is so much better when it comes to C++ compliance,
 which is very nice. However it's also a pain when you have to support
 lesser compilers like VC++ (man, it totally blows in the compliance
 department) and also previous gcc versions.
 
 Keep up the great work.
 -- 
 [ Below is a random fortune, which is unrelated to the above message. ]
 A bird in the bush usually has a friend in there with him.
 

Re: libstdc++/4083: input stream with field width() set reads too much to std::string

[Benjamin Kosnik]

The following reply was made to PR libstdc++/4083; it has been noted by GNATS.

From: Benjamin Kosnik <bkoz@redhat.com>
To: David Hedbor <david@caudium.net>
Cc: gcc-gnats@gcc.gnu.org
Subject: Re: libstdc++/4083: input stream with field width() set reads too much to std::string
Date: Fri, 24 Aug 2001 12:16:21 -0700 (PDT)

 > I assume that in the sections above, 21.3.7.9 is concerning
 > std::string and 27.6.1.2.3 is for char *?
 
 yep
 
 > Sorry for the incorrect bug report. 
 
 no worries. I'll close this one.

Re: libstdc++/4083: input stream with field width() set reads too much to std::string

[bkoz]

Synopsis: input stream with field width()  set reads too much to std::string

State-Changed-From-To: feedback->closed
State-Changed-By: bkoz
State-Changed-When: Fri Aug 24 12:19:10 2001
State-Changed-Why:
    Per request

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=4083&database=gcc

Re: libstdc++/4083: input stream with field width() set reads too much to std::string

[bkoz]

Synopsis: input stream with field width()  set reads too much to std::string

Responsible-Changed-From-To: unassigned->bkoz
Responsible-Changed-By: bkoz
Responsible-Changed-When: Fri Aug 24 12:07:52 2001
Responsible-Changed-Why:
    Mine.
State-Changed-From-To: open->feedback
State-Changed-By: bkoz
State-Changed-When: Fri Aug 24 12:07:52 2001
State-Changed-Why:
    This is not a bug. Both istream operators are working as expected:
    
    - 21.3.7.9 says that
    if is.width() is greater than 0, maximum number of appended chars is is.width, with characters extracted and appended until n characters are stored
    
    
    - 27.6.1.2.3 says that
    if is.width() is greater than 0, maximum number of appended chars is is.width, with characters extracted and appended until n-1 characters are stored
    
    Okay?

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=4083&database=gcc

libstdc++/4083: input stream with field width() set reads too much to std::string

[david]

>Number:         4083
>Category:       libstdc++
>Synopsis:       input stream with field width()  set reads too much to std::string
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Wed Aug 22 12:26:00 PDT 2001
>Closed-Date:
>Last-Modified:
>Originator:     David Hedbor
>Release:        3.0.1
>Organization:
>Environment:
Mandrake Cooker linux install, glibc 2.2.3
>Description:
When I set the width of an input stream and when use operator >> to an std::string, one byte too many is read. I.e
"ifs.width(8); ifs >> str;" reads 8 bytes if str is an std::string but 7 bytes if it's a char *. gcc 2.96 (mandrake) and egcs-2.91.66 do now show this problem.
>How-To-Repeat:
Compile and run the attached program.
>Fix:
Unknown. A workaround is to use a temporary char * for reading into.
>Release-Note:
>Audit-Trail:
>Unformatted:
----gnatsweb-attachment----
Content-Type: application/octet-stream; name="test.cpp"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="test.cpp"
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