Crouse wrote:
In the event of a sudden termination event for the program, it would
be nice to have a check there.
Look into "trap" and make sure no matter what your IFS gets set back
to it's original value
even if someone does a CTRL+ALT+DEL. ;)
Dave Crouse
Dave,
Thank you. I know trap. There is no need to reset IFS when run in a script
(see my comments in the script, "It's just good form". Why? When you execute a
script, it gets its own environment (a subshell so to speak) and nothing that
happens in a subshell can effect the current environment. For example, do this:
03:15 alchemy:~> echo "'$IFS'"
'
'
(That's ' \t\n' [space,tab,newline] by default)
Now let's change IFS in a subshell [ any command executed enclosed in
parenthesis on the command line gets its own subshell ], confirm the change,
then check the current environment:
22:44 alchemy:~> ( IFS='$__'; echo "'$IFS'" ); echo "'$IFS'"
'$__'
'
'
You can see that the IFS set and echoed in the subshell remained the $__ that
was assigned, but the second echo of IFS after the subshell terminated is still
the familiar old ' \t\n', That's when you don't need to reset it in a script,
but that is also why I do it anyway in the scripts and I write the comments in
the scripts telling you it isn't necessary but to do it anyway as a matter of
good form. Very few understand the subshell rule, but if you always reset IFS,
you will always be OK :p
Thanks for making me remember it though!
--
David C. Rankin, J.D.,P.E.
Rankin Law Firm, PLLC
510 Ochiltree Street
Nacogdoches, Texas 75961
Telephone: (936) 715-9333
Facsimile: (936) 715-9339
www.rankinlawfirm.com
