Difference between piped-in and invocation arguments to a Perl script?

pat_mc has asked for the wisdom of the Perl Monks concerning the following question:

Hi, Monks -

I came across a very, very basic observation today which illustrated to me that I am lacking some basic understanding on Perl's behaviour in a Unix-family operating system environment. My apologies therefore for bothering you with this question.

So far, I thought that piping some input into a Perl programme was the same as handing over that input as an invocation argument to the programme. I found, however, that this is not the case:

Here's my test programme test.pl:

#! /usr/bin/perl -w
print $ARGV[0];
[download]

My oberservations were:

$ test.pl Hello
Hello
[download]

while

echo "Hello" | test.pl
Use of uninitialized value in print at ./test.pl line 2
[download]

I thought that the pipe was passing on the arguments to the next element in the pipe via STDIN ... can anybody therefore please explain to me why this is not working?

Thanks for your help in advance!

Cheers - Pat

Comment on Difference between piped-in and invocation arguments to a Perl script? Select or Download Code

Replies are listed 'Best First'.
Re: Difference between piped-in and invocation arguments to a Perl script? by cdarke (Prior) on Jan 23, 2009 at 12:09 UTC
Passing a value as an argument (@ARGV) is not the same as reading from stdin, try: `#!/usr/bin/perl use warnings; use strict; my $value; if (defined $ARGV[0]) { $value = $ARGV[0]; } else { $value = <STDIN>; } print "<$value>\n";` [download] Run that using your two methods, and note the difference. Also lookup chomp. I hope that helps.	[reply] [d/l]
Re: Difference between piped-in and invocation arguments to a Perl script? by Corion (Patriarch) on Jan 23, 2009 at 12:18 UTC
In addition to the replies above, see Perl Documentation on I/O Operators, which explains what `<>` is, and how it relates to `@ARGV` and STDIN.	[reply] [d/l] [select]
Re: Difference between piped-in and invocation arguments to a Perl script? by targetsmart (Curate) on Jan 23, 2009 at 12:05 UTC
You have to use 'xargs' command for your requirement, for help put 'man xargs' in your shell. if you include xargs it will work echo "hello"\|xargs perl -e 'print $ARGV[0]'	[reply]
Re: Difference between piped-in and invocation arguments to a Perl script? by jrw (Monk) on Jan 23, 2009 at 17:25 UTC
Pat, actually your misunderstanding is at the unix level and has nothing to do with perl. Stdin is a totally different concept from command line arguments. The two are somewhat related by the standard use of filter programs which can be called like this: `filter_program FILE1 FILE2 generate_data \| filter_program` [download] The first form will read its input from the named files; the second form (with no command line arguments) will read its input from stdin. Both forms write their filtered output to stdout.	[reply] [d/l]
Re: Difference between piped-in and invocation arguments to a Perl script? by rekoil (Novice) on Jan 24, 2009 at 04:50 UTC
Simple but subtle = @ARGV != STDIN. The difference is that arguments are fed to the program at initialization time = they make up the values of the automatic @ARGV array - while STDIN is what is fed into a program after initialization via pipes, user input, or other redirects. So, if you want to pipe output to a perl app, use the <STDIN> filehandle to capture it, like so: `#!/usr/bin/perl -w my $line = <STDIN>; print "Line is: " . $line; $ echo 'echo' \| ./teststdin.pl Line is: echo` [download] This also works for redirecting files: `$ cat echo.txt echoed file $ ./teststdin.pl < echo.txt Line is: echoed file` [download] Hope this helps!	[reply] [d/l] [select]


Think about Loose Coupling
	PerlMonks