Perl as a Glue Language
=======================

(August 2016)

Here's a good header for our Perl files:

	#!/usr/bin/env perl

	use v5.10;
	use strict;
	use warnings;
	use autodie;

## Get command line arguments ##

	say scalar @ARGV;    # The number of arguments.
	say $ARGV[0];        # The first argument
	say $0;              # The name of this script.

## Reading a file line by line by line ##

	my $filename = '/tmp/stuff.txt';
	open my $fh, '<', $filename;
	while (my $line = <$fh>) {
		chomp $line;
		say $line;
	}

## Slurp in an entire (not too large) file at once ##

	my $filename = '/tmp/stuff.txt';
	open my $fh, '<', $filename;
	my $contents = do { local $/; <$fh> };

See an explanation at <http://www.perlmonks.org/?node_id=287647>.
In short, the built-in $/ tells Perl where to stop reading; it's set to newline by default.
We undefine $/ in this block scope in order to continue reading until the end of file.

## Check if a file exists ##

	say "Found the plain file $file_path" if -f $file_path

There are other file tests besides `-f`. See:

	$ perldoc -f '-X'

## Write to a file ##

	my $filename = '/tmp/stuff.txt';
	open my $fh, '>', $filename;
	print $fh "Here's a line of text\n";

To append rather than overwrite, open the filehandle with a double-carrot:

	open my $fh, '>>', $filename;

## Recurse through directory tree ##

	use File::Find;

	my @files;
	my $root = "/home/paulgorman/tmp";
	find(
		sub { push @files, $File::Find::name unless -d; },
		$root
	);
	for my $file (@files) {
		say $file;
	}

## All text files in a directory non-recursively ##

	my @files = glob ('/home/paulgorman/tmp/*.class');
	for my $file (@files) {
		say $file;
	}

## Print to STDERR ##

	print STDERR "We dun messed up.\n";

## Escape a string for the shell ##

Avoid the issue by using system():

	use IPC::System::Simple qw(system);
	system('command', 'arg1', 'arg2');

But if we must, look at `String::ShellQuote`.

## Run a shell command ##

## Get output of a shell command ##

## Hook up to the shell's standard handles ##


## Spawn a new process, and don't wait for it to exit ##


## Fork a Copy and Communicate via Unix Sockets ##

## Read from and Write to a Unix Named Pipe (FIFO) ##


## Web serve the current directory ##

$ cpan HTTP::Server::Brick
$ perl -MHTTP::Server::Brick -e \
	'$s=HTTP::Server::Brick->new(port=>8000); $s->mount("/"=>{path=>"."}); $s->start'

## Web serve a simple message ##


## Fetch URL (GET) ##

## Fetch URL (POST) ##

## Die! ##


## Links ##

- http://onyxneon.com/books/modern_perl/
- http://modernperlbooks.com/books/modern_perl_2016/