How to empty a file with Perl

Oct 26, 2016 by David Farrell

Have you ever had the experience of doing something a certain way for a long time, and then you discover a better way? This happened to me last week, when I was working on some code that needed to empty a file. Emptying a file is a common operation - maybe you have a session log file to write to, or want to limit disk space use, or whatever. Here’s how I usually do it:

# empty the file
{ open my $session_file, '>', 'path/to/sessionfile' }

This opens a write filehandle on the file, effectively setting its length to zero. I put the call to open between curly braces in order to minimize the scope of the filehandle, $session_file. After that statement, the block closes, and $session_file variable goes out of scope, automatically closing the filehandle. As the block looks a little strange, I include an explanatory comment.

The other day though, I came across the truncate function. This does exactly what you’d think it does: truncates files. It takes two arguments: the file path (or filehandle), and the length. So if you need to truncate a file, you can do:

truncate 'path/to/sessionfile', 0;

This doesn’t use a lexical variable, so no scoping is required. It’s unambiguous so no comment is needed either. I like it, it’s a better way.

N.B. on Windows truncate requires the file to not be open elsewhere on the system, and if called with a filehandle it must be in append mode.

Looking up Perl functions

Do you know Perl has around 220 built in functions? You can read about them all in perlfunc, or at the terminal with perldoc perlfunc. Read more about the truncate function at the terminal with perldoc -f truncate.

Vim users if you’re editing Perl code and want to lookup a function, place the cursor on the function word and type Shift-k to lookup the function in perldoc (works for Python, Ruby, C etc too).

This article was originally posted on PerlTricks.com.