From blgl@hagernas.com

Created by blgl@hagernas.com

Truncating a string isn't an expensive operation,
so is this really too much to ask?

#! /usr/local/bin/perl
use strict;
open(my $fh,'+>',\my $buf)
  or die "open​: $!\n";
print {$fh} "Hello, World!\n"
  or die "print​: $!\n";
truncate($fh,5)
  or warn "truncate​: $!\n"; # outputs "truncate​: Bad file
descriptor"
print 'length​: '.length($buf)."\n"; # outputs "length​: 14"

Flags:
     category=core
     severity=low

Site configuration information for perl v5.8.8:

Configured by blgl at Fri Jun  9 19:26:53 CEST 2006.

Summary of my perl5 (revision 5 version 8 subversion 8) configuration:
   Platform:
     osname=darwin, osvers=8.6.0, archname=darwin-thread- 
multi-64int-2level
     uname='darwin sunwukung.sth.ownit.se 8.6.0 darwin kernel version  
8.6.0: tue mar 7 16:58:48 pst 2006;  
root:xnu-792.6.70.obj~1release_ppc power macintosh powerpc '
     config_args=''
     hint=recommended, useposix=true, d_sigaction=define
     usethreads=define use5005threads=undef useithreads=define  
usemultiplicity=define
     useperlio=define d_sfio=undef uselargefiles=define usesocks=undef
     use64bitint=define use64bitall=undef uselongdouble=undef
     usemymalloc=n, bincompat5005=undef
   Compiler:
     cc='cc', ccflags ='-fno-common -DPERL_DARWIN -no-cpp-precomp - 
fno-strict-aliasing -pipe -Wdeclaration-after-statement -I/usr/local/ 
include',
     optimize='-Os',
     cppflags='-no-cpp-precomp -fno-common -DPERL_DARWIN -no-cpp- 
precomp -fno-strict-aliasing -pipe -Wdeclaration-after-statement -I/ 
usr/local/include'
     ccversion='', gccversion='4.0.1 (Apple Computer, Inc. build  
5247)', gccosandvers=''
     intsize=4, longsize=4, ptrsize=4, doublesize=8, byteorder=87654321
     d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=16
     ivtype='long long', ivsize=8, nvtype='double', nvsize=8,  
Off_t='off_t', lseeksize=8
     alignbytes=8, prototype=define
   Linker and Libraries:
     ld='env MACOSX_DEPLOYMENT_TARGET=10.3 cc', ldflags =' -L/usr/ 
local/lib'
     libpth=/usr/local/lib /usr/lib
     libs=-lc
     perllibs=-lc
     libc=/usr/lib/libc.dylib, so=dylib, useshrplib=false,  
libperl=libperl.a
     gnulibc_version=''
   Dynamic Linking:
     dlsrc=dl_dlopen.xs, dlext=bundle, d_dlsymun=undef, ccdlflags=' '
     cccdlflags=' ', lddlflags=' -bundle -undefined dynamic_lookup -L/ 
usr/local/lib'

Locally applied patches:



@INC for perl v5.8.8:
     /usr/local/lib/perl5/5.8.8/darwin-thread-multi-64int-2level
     /usr/local/lib/perl5/5.8.8
     /usr/local/lib/perl5/site_perl/5.8.8/darwin-thread- 
multi-64int-2level
     /usr/local/lib/perl5/site_perl/5.8.8
     /usr/local/lib/perl5/site_perl
     .


Environment for perl v5.8.8:
     DYLD_LIBRARY_PATH (unset)
     HOME=/Users/blgl
     LANG (unset)
     LANGUAGE (unset)
     LD_LIBRARY_PATH (unset)
     LOGDIR (unset)
     PATH=/Users/blgl/bin:/Users/blgl/bin/powerpc-apple-darwin:/usr/ 
local/bin:/usr/local/sbin:/bin:/sbin:/usr/bin:/usr/sbin
     PERL_BADLANG (unset)
     SHELL=/bin/tcsh

From @rgarcia

On 26/08/06, via RT Bo Lindbergh <perlbug-followup@​perl.org> wrote​:

Truncating a string isn't an expensive operation,
so is this really too much to ask?

A quick look at the implementation suggests that truncate() only works
for filehandles currently.

The RT System itself - Status changed from 'new' to 'open'

From nick@ing-simmons.net

Rafael Garcia-Suarez <rgarciasuarez@​gmail.com> writes​:

On 26/08/06, via RT Bo Lindbergh <perlbug-followup@​perl.org> wrote​:

Truncating a string isn't an expensive operation,
so is this really too much to ask?

A quick look at the implementation suggests that truncate() only works
for filehandles currently.

But a PerlIO​::Scalar "is" a file handle :-(

That said we there isn't a ftruncate() hook in PerlIO virtualization
so I don't know what we do...

From blgl@hagernas.com

In article <20060901184007.24952.7@​llama.ing-simmons.net>,
nick@​ing-simmons.net (Nick Ing-Simmons) wrote​:

Rafael Garcia-Suarez <rgarciasuarez@​gmail.com> writes​:

On 26/08/06, via RT Bo Lindbergh <perlbug-followup@​perl.org> wrote​:

Truncating a string isn't an expensive operation,
so is this really too much to ask?

A quick look at the implementation suggests that truncate() only works
for filehandles currently.

But a PerlIO​::Scalar "is" a file handle :-(

That said we there isn't a ftruncate() hook in PerlIO virtualization
so I don't know what we do...

Adding _only_ ftruncate to PerlIO is not the Right Thing to do,
because the next bug report would be "Why can't you use the -s
operator to get the length of the scalar used by a
PerlIO​::Scalar handle?" And once you have both fstat and
ftruncate, you might as well complete the set and add fchown,
fchmod, and futimes too.

/Bo Lindbergh

From nick@ing-simmons.net

Bo Lindbergh <blgl@​hagernas.com> writes​:

In article <20060901184007.24952.7@​llama.ing-simmons.net>,
nick@​ing-simmons.net (Nick Ing-Simmons) wrote​:

Rafael Garcia-Suarez <rgarciasuarez@​gmail.com> writes​:

On 26/08/06, via RT Bo Lindbergh <perlbug-followup@​perl.org> wrote​:

Truncating a string isn't an expensive operation,
so is this really too much to ask?

A quick look at the implementation suggests that truncate() only works
for filehandles currently.

But a PerlIO​::Scalar "is" a file handle :-(

That said we there isn't a ftruncate() hook in PerlIO virtualization
so I don't know what we do...

Adding _only_ ftruncate to PerlIO is not the Right Thing to do,
because the next bug report would be "Why can't you use the -s
operator to get the length of the scalar used by a
PerlIO​::Scalar handle?"

And I would agree with that.
But then we have a bunch of other fields in stat that we would have to
lie about or fake by tracking last time touched etc.

And once you have both fstat and
ftruncate, you might as well complete the set and add fchown,
fchmod, and futimes too.

If we fake times in our fake stat then fake futimes makes sense.
But chown an chmod make little sense to me - only the current process
can access the scalar so current UID is owner (and can't exec it).

We could accept the calls and set errno to nearest code that made sense,
or perhaps just set the mode so that fake stat saw it.

Have we missed any more fd related calls?
If there anything else that we can do to filehandles via perl
interface that we missed?

From blgl@hagernas.com

In article <20060902095359.24952.8@​llama.ing-simmons.net>,
nick@​ing-simmons.net (Nick Ing-Simmons) wrote​:

If we fake times in our fake stat then fake futimes makes sense.
But chown an chmod make little sense to me - only the current process
can access the scalar so current UID is owner (and can't exec it).

That's true for scalar i/o, but they would be useful for other kinds
of virtual file. An SFTP client comes to mind.

Have we missed any more fd related calls?

ioctl and fcntl. I don't know about the former, but a virtual
O_NONBLOCK flag might be useful.

/Bo Lindbergh

From @leonerd

I find this bug still present, and affecting a unit test of mine
(wherein I construct scalar filehandles to pass to the function
under test to avoid touching the real FS)​:

  $ perlsh
  eval​: open my $fh, ">", \my $content;
  '1'

  eval​: $fh->print( "AAAAAAAAAA" );
  '1'

  eval​: $fh->seek(0,0); $fh->truncate(0)
  undef

  eval​: $fh->print( "BB" )
  '1'

  eval​: $content
  'BBAAAAAAAA'

This is on​:

$ perl -v

This is perl 5, version 18, subversion 1 (v5.18.1) built for
x86_64-linux-gnu-thread-multi(with 46 registered patches, see perl -V
for more detail)

While I await a fix - any known workarounds? E.g. a Tie module?

--
Paul "LeoNerd" Evans

leonerd@​leonerd.org.uk
ICQ# 4135350 | Registered Linux# 179460
http​://www.leonerd.org.uk/