gnu@perl has asked for the wisdom of the Perl Monks concerning the following question:

I have a program that when it starts tries to obtain an exclusive lock on a file. In order to do that I have to open the file allowing write. I have done it like this open(FH,"+<./program.pid"); since I can't have the file clobbered.

I now need to request an exclusive lock on that file. If granted, I know that no other instance of this program is running and I can continue. Now I need to read this file, it contains the exit status of the last instance of this program. This lets me know where I need to start up at (I maintain a state file).

After this file is read and I know where to begin in my processing I wish to truncate this file and add the information from this instance of the program. Can I just reopen the file like this open(FH,">./program.pid");? Will this just replace the previous instance of 'FH' or will I be stacking things up?

Also a concern, when I reopen the file with the same filehandle, it is (I think) essentially closing the previous one. This would make me lose my exclusive lock on the file. I could immediately place an exclusive lock again, but this could end up an a race state with another version of this program.

Here's a little example of what I want to do:

... open(PIDFILE,">>$pidFile") or die "Cannot open pidfile, $pidFile: $!\n"; if (! &lock(*PIDFILE) ) { print "\n\tCannot obtain exclusive lock on $pidFile\n", "\tThere must be another instance of this\n", "\tprogram running. Exiting.\n\n"; exit; } # GOT LOCK my @arr = (<PIDFILE>); open(PIDFILE,">$pidFile"); print PIDFILE "$$\n"; ...

Replies are listed 'Best First'.
Re: Reopening a file handle
by gnu@perl (Pilgrim) on Mar 04, 2003 at 21:41 UTC
    UPDATE: After some more research and searching the web I discovered the 'truncate' command. What a wonderful find, you would think I would have run into this long ago, but I guess you keep learning.

    After finding that the only other thing I had to do was autoflush the 'FH'.