On Wed, Jan 30, 2002 at 08:24:11PM +0200, Jarkko Hietaniemi wrote:
>  On Wed, Jan 30, 2002 at 04:02:01PM +0100, Arthur Bergman wrote:

>  > 2) libs/warning
>  > error with regard to files opened in a certain mode, should we skip this test since we k
now it is failing?
>  
>  I never understood why are we still getting this error-- I thought
>  it was bound to broken/different implementation of *stdio* in certain
>  platforms, and now in the shiny new era of perlio...?  Nick?

You were meaning Nick I-S, but I'll express an opinion
>  In case someone is wondering, this the pp_hot warning:


>  ########
>  # pp_hot.c [pp_print]
              ^^^^^^^^
we're testing this code with the prints to STDIN, and something completely
different with the <> 

>  use warnings 'io' ;
>  print STDIN "anc";
>  print <STDOUT>;
>  print <STDERR>;
>  open(FOO, ">&STDOUT") and print <FOO>;
>  print getc(STDERR);
>  print getc(FOO);

code which has no references to STD anything

>  ####################################################################
>  # The next test is known to fail on some systems (Linux+old glibc, #
>  # some *BSDs (including Mac OS X and NeXT), among others.          #
>  # We skip it for now (on the grounds that it is "just" a warning). #
>  ####################################################################
>  #read(FOO,$_,1);
>  no warnings 'io' ;
>  print STDIN "anc";
>  EXPECT
>  Filehandle STDIN opened only for input at - line 3.
>  Filehandle STDOUT opened only for output at - line 4.
>  Filehandle STDERR opened only for output at - line 5.
>  Filehandle FOO opened only for output at - line 6.
>  Filehandle STDERR opened only for output at - line 7.
>  Filehandle FOO opened only for output at - line 8.

And it transpires that STDOUT is read/write on FreeBSD (and maybe some other
places) so the code as written is an unintended bit of non-portability.

I vote that we re-write it to open a temporary file for input, and then
try to print to that, and remove all reference to STDIN.
Split out the <>  reading tests to another test, which I think is for
do_readline (which is actually in pp_hot.c)

If this is not thought to be a bad way I'll write a patch.
And "live dangerously" and reinstate the skipped read which I believe is
due to expecting STDOUT to be write only.

Nicholas Clark
-- 
EMCFT http://www.ccl4.org/~nick/CV.html