Cover image generated using carbon

This post will give an overview of perl syntax for command line usage and some examples to show what kind of problems are typically suited for one-liners.

Why use Perl for one-liners?

I assume you are already familiar with use cases where command line is more productive compared to GUI. See also this series of articles titled Unix as IDE.

A shell utility like bash provides built-in commands and scripting features to easily solve and automate various tasks. External *nix commands like grep, sed, awk, sort, find, parallel, etc can be combined to work with each other. Depending upon your familiarity with those tools, you can either use perl as a single replacement or complement them for specific use cases.

Here's some one-liners (options will be explained later):

• perl -pe 's/(?:\x27;\x27|";")(*SKIP)(*F)|;/#/g' — change ; to # but don't change ; within single or double quotes
• perl -MList::Util=uniq -e 'print uniq <>' — retain only first copy of duplicated lines, uses built-in module List::Util
• perl -MRegexp::Common=net -nE 'say join "\n", //g if /$RE{net}{IPv4}/' — extract only IPv4 addresses, using a third-party Regexp::Common module
• Some stackoverflow Q&A that I've answered over the years with simpler perl solution compared to other cli tools
  • replace string with incrementing value
  • sort rows in csv file without header & first column
  • reverse matched pattern
  • append zeros to list
  • arithmetic replacement in a text file
  • reverse complement DNA sequence for a specific field

The selling point of perl over tools like grep, sed and awk includes feature rich regular expression engine and standard/third-party modules. If you don't already know the syntax and idioms for sed and awk, learning command line options for perl would be the easier option. Another advantage is that perl is more portable compared to GNU, BSD and Mac implementations of cli tools. The main disadvantage is that perl is likely to be slower for features that are supported out of the box by those tools.

See also unix.stackexchange: when to use grep, sed, awk, perl, etc

Command line options

perl -h gives the list of all command line options, along with a brief description. See perldoc: perlrun for documentation on these command switches.

This post will show examples with -e, -l, -n, -p and -a options.

Executing Perl code

If you want to execute a perl program file, one way is to pass the filename as argument to the perl command.

$ echo 'print "Hello Perl\n"' > hello.pl
$ perl hello.pl
Hello Perl

For short programs, you can also directly pass the code as an argument to the -e or -E options. See perldoc: feature for details about the features enabled by the -E option.

$ perl -e 'print "Hello Perl\n"'
Hello Perl

$ # multiple statements can be issued separated by ;
$ # -l option will be covered in detail later, appends \n to 'print' here
$ perl -le '$x=25; $y=12; print $x**$y'
59604644775390625
$ # or, use -E and 'say' instead of -l and 'print'
$ perl -E '$x=25; $y=12; say $x**$y'
59604644775390625

Filtering

perl one-liners can be used for filtering lines matched by a regexp, similar to grep, sed and awk. And similar to many command line utilities, perl can accept input from both stdin and file arguments.

$ # sample stdin data
$ printf 'gate\napple\nwhat\nkite\n'
gate
apple
what
kite

$ # print all lines containing 'at'
$ # same as: grep 'at' and sed -n '/at/p' and awk '/at/'
$ printf 'gate\napple\nwhat\nkite\n' | perl -ne 'print if /at/'
gate
what

$ # print all lines NOT containing 'e'
$ # same as: grep -v 'e' and sed -n '/e/!p' and awk '!/e/'
$ printf 'gate\napple\nwhat\nkite\n' | perl -ne 'print if !/e/'
what

By default, grep, sed and awk will automatically loop over input content line by line (with \n as the line distinguishing character). The -n or -p option will enable this feature for perl.

As seen before, the -e option accepts code as command line argument. Many shortcuts are available to reduce the amount of typing needed. In the above examples, a regular expression (defined by the pattern between a pair of forward slashes) has been used to filter the input. When the input string isn't specified, the test is performed against special variable $_, which has the contents of the current input line here (the correct term would be input record). $_ is also the default argument for many functions like print and say. To summarize:

• /REGEXP/FLAGS is a shortcut for $_ =~ m/REGEXP/FLAGS
• !/REGEXP/FLAGS is a shortcut for $_ !~ m/REGEXP/FLAGS

See perldoc: match for help on m operator. See perldoc: special variables for documentation on $_, $&, etc.

Here's an example with file input instead of stdin.

$ cat table.txt
brown bread mat hair 42
blue cake mug shirt -7
yellow banana window shoes 3.14

$ perl -nE 'say $& if /(?<!-)\d+$/' table.txt
42
14

$ # if the condition isn't required, capture groups can be used
$ perl -nE 'say /(\d+)$/' table.txt
42
7
14

Substitution

Use s operator for search and replace requirements. By default, this operates on $_ when the input string isn't provided. For these examples, -p option is used instead of -n option, so that the value of $_ is automatically printed after processing each input line. See perldoc: search and replace for documentation and examples.

$ # for each input line, change only first ':' to '-'
$ # same as: sed 's/:/-/' and awk '{sub(/:/, "-")} 1'
$ printf '1:2:3:4\na:b:c:d\n' | perl -pe 's/:/-/'
1-2:3:4
a-b:c:d

$ # for each input line, change all ':' to '-'
$ # same as: sed 's/:/-/g' and awk '{gsub(/:/, "-")} 1'
$ printf '1:2:3:4\na:b:c:d\n' | perl -pe 's/:/-/g'
1-2-3-4
a-b-c-d

The s operator modifies the input string it is acting upon if the pattern matches. In addition, it will return number of substitutions made if successful, otherwise returns a false value (empty string or 0). You can use r flag to return string after substitution instead of in-place modification. See perldoc: perlretut for a tutorial on Perl regular expressions.

Field processing

Consider the sample input file shown below with fields separated by a single space character.

$ cat table.txt
brown bread mat hair 42
blue cake mug shirt -7
yellow banana window shoes 3.14

Here's some examples that is based on specific field rather than the entire line. The -a option will cause the input line to be split based on whitespaces and the field contents can be accessed using @F special array variable. Leading and trailing whitespaces will be suppressed, so there's no possibility of empty fields.

$ # print the second field of each input line
$ # same as: awk '{print $2}' table.txt
$ perl -lane 'print $F[1]' table.txt
bread
cake
banana

$ # print lines only if last field is a negative number
$ # same as: awk '$NF<0' table.txt
$ perl -lane 'print if $F[-1] < 0' table.txt
blue cake mug shirt -7

$ # change 'b' to 'B' only for the first field
$ # same as: awk '{gsub(/b/, "B", $1)} 1' table.txt
$ perl -lane '$F[0] =~ s/b/B/g; print "@F"' table.txt
Brown bread mat hair 42
Blue cake mug shirt -7
yellow banana window shoes 3.14

BEGIN and END

You can use a BEGIN{} block when you need to execute something before input is read and a END{} block to execute something after all of the input has been processed.

$ # same as: awk 'BEGIN{print "---"} 1; END{print "%%%"}'
$ seq 4 | perl -pE 'BEGIN{say "---"} END{say "%%%"}'
---
1
2
3
4
%%%

ENV hash

When it comes to automation and scripting, you'd often need to construct commands that can accept input from user, file, output of a shell command, etc. As mentioned before, this book assumes bash as the shell being used. To access environment variables of the shell, you can use the special hash variable %ENV with the name of the environment variable as a string key.

$ # existing environment variable
$ # output shown here is for my machine, would differ for you
$ perl -E 'say $ENV{HOME}'
/home/learnbyexample
$ perl -E 'say $ENV{SHELL}'
/bin/bash

$ # defined along with perl command
$ # note that the variable is placed before the shell command
$ word='hello' perl -E 'say $ENV{word}'
hello
$ # the input characters are preserved as is
$ ip='hi\nbye' perl -E 'say $ENV{ip}'
hi\nbye

Here's another example when a regexp is passed as an environment variable content.

$ cat word_anchors.txt
sub par
spar
apparent effort
two spare computers
cart part tart mart

$ # assume 'r' is a shell variable that has to be passed to the perl command
$ r='\Bpar\B'
$ rgx="$r" perl -ne 'print if /$ENV{rgx}/' word_anchors.txt
apparent effort
two spare computers

You can also make use of the -s option to assign a perl variable.

$ r='\Bpar\B'
$ perl -sne 'print if /$rgx/' -- -rgx="$r" word_anchors.txt
apparent effort
two spare computers

As an example, see my repo ch: command help for a practical shell script, where commands are constructed dynamically.

Executing external commands

You can execute external commands using the system function. See perldoc: system for documentation and details like how string/list argument is processed before it is executed.

$ perl -e 'system("echo Hello World")'
Hello World

$ perl -e 'system("wc -w <word_anchors.txt")'
12

$ perl -e 'system("seq -s, 10 > out.txt")'
$ cat out.txt
1,2,3,4,5,6,7,8,9,10

Return value of system or special variable $? can be used to act upon exit status of command issued. As per documentation:

The return value is the exit status of the program as returned by the wait call. To get the actual exit value, shift right by eight

$ perl -E '$es=system("ls word_anchors.txt"); say $es'
word_anchors.txt
0
$ perl -E 'system("ls word_anchors.txt"); say $?'
word_anchors.txt
0

$ perl -E 'system("ls xyz.txt"); say $?'
ls: cannot access 'xyz.txt': No such file or directory
512

To save the result of an external command, use backticks or qx operator. See perldoc: qx for documentation and details like separating out STDOUT and STDERR.

$ perl -e '$words = `wc -w <word_anchors.txt`; print $words'
12

$ perl -e '$nums = qx/seq 3/; print $nums'
1
2
3

See also stackoverflow: difference between backticks, system, and exec

Summary

This post introduced some of the common options for perl cli usage, along with typical cli text processing examples. While specific purpose cli tools like grep, sed and awk are usually faster, perl has a much more extensive standard library and ecosystem. And you do not have to learn a lot if you are already comfortable with perl but not familiar with those cli tools.

Perl one-liners cookbook

If you liked this post and would like to learn more, check out my ebook using the links below.

• https://gumroad.com/l/perl-oneliners
• https://leanpub.com/perl-oneliners

You can also get the ebooks as part of Magical one-liners bundle using these links:

• https://gumroad.com/l/oneliners
• https://leanpub.com/b/oneliners