Scroll to navigation

IPC::Run::Debug(3) User Contributed Perl Documentation IPC::Run::Debug(3)

NAME

IPC::Run::Debug - debugging routines for IPC::Run

SYNOPSIS

   ##
   ## Environment variable usage
   ##
   ## To force debugging off and shave a bit of CPU and memory
   ## by compile-time optimizing away all debugging code in IPC::Run
   ## (debug => ...) options to IPC::Run will be ignored.
   export IPCRUNDEBUG=none
   ## To force debugging on (levels are from 0..10)
   export IPCRUNDEBUG=basic
   ## Leave unset or set to "" to compile in debugging support and
   ## allow runtime control of it using the debug option.

DESCRIPTION

Controls IPC::Run debugging. Debugging levels are now set by using words, but the numbers shown are still supported for backwards compatibility:

   0  none         disabled (special, see below)
   1  basic        what's running
   2  data         what's being sent/received
   3  details      what's going on in more detail
   4  gory         way too much detail for most uses
   10 all          use this when submitting bug reports
      noopts       optimizations forbidden due to inherited STDIN

The "none" level is special when the environment variable IPCRUNDEBUG is set to this the first time IPC::Run::Debug is loaded: it prevents the debugging code from being compiled in to the remaining IPC::Run modules, saving a bit of cpu.

To do this in a script, here's a way that allows it to be overridden:

   BEGIN {
      unless ( defined $ENV{IPCRUNDEBUG} ) {
         eval 'local $ENV{IPCRUNDEBUG} = "none"; require IPC::Run::Debug"'
            or die $@;
      }
   }

This should force IPC::Run to not be debuggable unless somebody sets the IPCRUNDEBUG flag; modify this formula to grep @ARGV if need be:

   BEGIN {
      unless ( grep /^--debug/, @ARGV ) {
         eval 'local $ENV{IPCRUNDEBUG} = "none"; require IPC::Run::Debug"'
         or die $@;
   }

Both of those are untested.

AUTHOR

Barrie Slaymaker <barries@slaysys.com>, with numerous suggestions by p5p.

2021-08-09 perl v5.32.1