close for piped open sets $! to the wrong value

[p5pRT]

Migrated from rt.perl.org#2113 (status was 'resolved')

Searchable as RT2113$

[p5pRT]

From @mjdominus

Created by @mjdominus

This program runs /bin/false\, a standard unix command\, on a piped open\, then closes the filehandle. /bin/false is the program which exits immediately with exit status 1. According to the manual​:

  If the file handle came from a piped open C\<close()> will   additionally return FALSE if one of the other system calls   involved fails or if the program exits with non-zero status.   (If the only problem was that the program exited non-zero   C\<$!> will be set to C\<0>.)

  [perlfunc/close]

Since /bin/false exited with non-zero status\, $! should be set to 0. However\, it is instead set to the empty string.

  #!/usr/bin/perl   open FH\, "/bin/false |" or die "fork​: $!";   print ">> $_" while \;   my $rc = close FH;   my $errstr = defined($!) ? $! : "undefined";   $rc = defined($rc) ? $rc : "undefined";   print "close rc​: ($rc); \$!​: ($errstr); \$?​: ($?)\n";

The output is​:

  close rc​: (); $!​: (); $?​: (256)

If there is any version of Perl that has ever set $! to 0\, I cannot find it. Versions up through 5.003 set $! to `Illegal seek' (which I guess is probably system-dependent) and versions 5.004 and later show the behavior above.

 

Perl Info

``` Site configuration information for perl 5.00503: Configured by mjd at Sun Jan 23 14:15:22 EST 2000. Summary of my perl5 (5.0 patchlevel 5 subversion 3) configuration: Platform: osname=linux, osvers=2.2.12, archname=i586-linux uname='linux plover 2.2.12 #10 mon nov 1 17:02:09 est 1999 i586 unknown ' hint=recommended, useposix=true, d_sigaction=define usethreads=undef useperlio=undef d_sfio=undef Compiler: cc='cc', optimize='-O2', gccversion=egcs-2.90.29 980515 (egcs-1.0.3 release) cppflags='-Dbool=char -DHAS_BOOL -I/usr/local/include' ccflags ='-Dbool=char -DHAS_BOOL -I/usr/local/include' stdchar='char', d_stdstdio=define, usevfork=false intsize=4, longsize=4, ptrsize=4, doublesize=8 d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=12 alignbytes=4, usemymalloc=n, prototype=define Linker and Libraries: ld='cc', ldflags =' -L/usr/local/lib' libpth=/usr/local/lib /shlib /lib /usr/lib libs=-lndbm -lgdbm -ldbm -ldb -ldl -lm -lc libc=/lib/libc.so.6, so=so, useshrplib=false, libperl=libperl.a Dynamic Linking: dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags='-rdynamic' cccdlflags='-fpic', lddlflags='-shared -L/usr/local/lib' Locally applied patches: @INC for perl 5.00503: /usr/lib/perl5/5.00503/i586-linux /usr/lib/perl5/5.00503 /usr/lib/perl5/site_perl/5.005/i586-linux /usr/lib/perl5/site_perl/5.005 . Environment for perl 5.00503: HOME=/home/mjd LANG (unset) LANGUAGE (unset) LD_LIBRARY_PATH=/lib:/usr/lib:/usr/X11R6/lib LOGDIR (unset) PATH=/home/mjd/bin:/usr/local/bin:/bin:/usr/bin:/usr/X11/bin:/usr/games:/sbin:/usr/sbin:/usr/local/bin/X11:/usr/local/bin/mh:/data/mysql/bin:/home/mjd/TPI/bin:/usr/local/mysql/bin PERL_BADLANG (unset) SHELL=/bin/bash ```

[p5pRT]

From [Unknown Contact. See original ticket]

mjd@​plover.com writes​:

Since /bin/false exited with non-zero status\, $! should be set to 0. However\, it is instead set to the empty string.

Do you remember that $! is double-headed? It has both numeric and string values...

Ilya

[p5pRT]

From @mjdominus

mjd@​plover.com writes​:

Since /bin/false exited with non-zero status\, $! should be set to 0. However\, it is instead set to the empty string.

Do you remember that $! is double-headed? It has both numeric and string values...

Maybe the documentation is at fault\, then. I don't see why the current behavior is not simply described by

  If the only problem was that the program exited non-zero $!   will be set to the empty string.

[p5pRT]

From [Unknown Contact. See original ticket]

On Mon\, Feb 07\, 2000 at 04​:19​:54PM -0500\, Mark-Jason Dominus wrote​:

Do you remember that $! is double-headed? It has both numeric and string values...

Maybe the documentation is at fault\, then. I don't see why the current behavior is not simply described by

If the only problem was that the program exited non\-zero $\!
will be set to the empty string\.

Because the value of $! *is not a string*.

Ilya

[p5pRT]

From @mjdominus

Because the value of $! *is not a string*.

If it were a string\, how would the behvaior be different in this case?

[p5pRT]

From [Unknown Contact. See original ticket]

Maybe the documentation is at fault\, then. I don't see why the current behavior is not simply described by

If the only problem was that the program exited non\-zero $\!
will be set to the empty string\.

Because the value of $! *is not a string*.

Wrong. Back into the Sophist's hole with you. Just looking for crap to pick at today\, eh?

It can be a string. It can be also a number.

And these are magically rather than canonically connected.

So what.

--tom