STDIN and IPC::Open3 -- 20 paces. pistols. noon

[Eric Wilhelm]

# from Michael G Schwern
# on Tuesday 11 September 2007 03:16 pm:

>That's testing whether STDIN is a TTY.  If TAP::Parser is doing
> anything with STDIN then its hosed.  That appears to be the problem.
>  Here's a cut down test.
>
>  #!/usr/bin/perl -w
>
>  use Test::More tests => 1;
>
>  ok -t STDIN;

I can't find a good way through this one.  We need IPC::OpenTheOtherTwo 
or thereabouts because Open3 is doing this:

    # If the write handle is a dup give it away entirely, close my copy
    # of it.
    xclose $dad_wtr if $dup_wtr;

Which, if called with '<&STDIN'[1], closes STDIN.  So, it will work 
once, but not twice:

  $ runtests t/stdin.t
  # yay

  $ runtests t/000-load.t t/stdin.t
  # explosions, blood, gore, tears

Thus (AFAICT), to get our STDIN to the child while retaining control of 
STDOUT and STDERR, we have to patch Open3.

[1] there are other \*STDIN and various other possibilities, but I can't 
see any way to make open3 not close it.  Also note the third 
combination may yield a different answer depending on the planetary 
alignments:

  $ runtests t/stdin.t t/000-load.t
  # gore, explosions, tears, blood

--Eric
-- 
But you can never get 3n from n, ever, and if you think you can, please
email me the stock ticker of your company so I can short it.
--Joel Spolsky
---------------------------------------------------
    http://scratchcomputing.com
---------------------------------------------------