One-liners

Reverse every line:

Input file:

$ cat foo
qwe
123
bar

Reverse file lines:

$ ruby -e 'File.open("foo").each_line { |l| puts l.chop.reverse }'
ewq
321
rab

Reverse lines from stdout:

cat foo | ruby -e 'while s = gets; puts s.chop.reverse; end'

Reverse both (ARGF == $<):

cat foo | ruby -e 'ARGF.each_line { |l| puts l.chop.reverse }'
ruby -e 'ARGF.each_line { |l| puts l.chop.reverse }' foo

Let Ruby iterate for you:

ruby -ne 'puts $_.chop.reverse' foo

Let Ruby print for you (clumsy):

ruby -pe '$_.chop!.reverse!; $_ += "\n"' foo

Let Ruby chop for you:

ruby -lpe '$_.reverse!' foo

In-place file editing:

$ cat foo
qwe
123
bar
$ ruby -i -lpe '$_.reverse!' foo
$ cat foo
ewq
321
rab

Indent every line:

Just that:

cat example.rb | ruby -ne 'puts " " * 4 + $_'

Or even that:

ruby -ne 'puts " " * 4 + $_' example.rb

Line numbering (all files at once):

ruby -ne 'puts "#$. #$_"' foo.rb bar.rb

Line numbering (per file file):

ruby -ne '$. = 1 if $<.pos - $_.size == 0; puts "#$. #$_"' foo.rb bar.rb

Count words:

ruby -ane 'w = (w || 0) + $F.size; END { p w }' exmpl.txt

Delete all consecutive blank lines from file except the first one in each group:

ruby -ne 'puts $_ if /^[^\n]/../^$/'

Extract all comments:

Input:

$ cat comments.txt 
some code
/*
  comment one
  comment two
*/
more code
/* one-line comment */
a bit more code
/*
  comment three
*/
even more code

Ruby in action:

$ ruby -ne 'puts $_ if ($_ =~ /^\/\*/)..($_ =~ /\*\/$/)' comments.txt
/*
  comment one
  comment two
*/
/* one-line comment */
/*
  comment three
*/

Alternative:

ruby -pe 'next unless ($_ =~ /^\/\*/)..($_ =~ /\*\/$/)' comments.txt

Homebrew flip-flop (from "The Ruby programming language" by David Flanagan, Yukihiro Matsumoto):

# flip-flop.rb
$state = false
def flipflop(s)
  unless $state
    result = (s =~ /^\/\*$/)
    if result
      $state = !(s =~ /^\*\/$/)
    end 
    result
  else
    $state = !(s =~ /^\*\/$/)
    true
  end 
end

ARGF.each_line do |l| 
  puts l if flipflop(l)
end

Try it:

ruby flip-flop.rb comments.txt

Highlight trailing spaces:

ruby -lpe '$_.gsub! /(\s+)$/, "\e[41m\\1\e[0m"' example.rb

Remove trailing spaces:

ruby -lpe '$_.rstrip!' example.rb

For each line in the file highlights with red the part of it that goes over 50 characters:

Easy way:

ruby -ne 'puts "#{$_}\e[31m#{$_.chop!.slice!(60..-1)}\e[0m"' example.rb

Uglier, but configurable:

ruby -e 'w = $*.shift; $<.each { |l| puts "#{l}\e[31m#{l.chop!.slice!(w.to_i..-1)}\e[0m" }' 50 example.rb

Simplest REPL:

Straightforward one:

ruby -e  'loop { puts eval(gets) }'

Leverage -n option:

ruby -ne 'puts eval($_)'

Prompt (but not in the 1st line):

ruby -ne 'print "#{eval($_).inspect}\n>> "'

Fix it:

ruby -ne 'BEGIN{print">> "}; print "#{eval($_).inspect}\n>> "'

REPL should be polite:

ruby -ne 'BEGIN{print">> "}; print "#{eval($_).inspect}\n>> "; END{puts "Bye"}'

Handle RuntimeError (e.g., NameError), but not SyntaxError:

ruby -ne 'BEGIN{print">> "}; print "#{(eval($_) rescue $!).inspect}\n>> "; END{puts "Bye"}'

Rescue everything - save the world:

ruby -ne 'BEGIN{print">> "}; print "#{(begin; eval($_); rescue Exception; $!; end).inspect}\n>> "; END{puts "Bye"}'

Additions

# copyright.rb
copyright = DATA.read

ARGF.each_line do |l|
  puts copyright if $. == 1
  puts l
end

__END__
####################################
# Copyright (C) 2012 Kirill Lashuk #
####################################

Use it:

ruby copyright.rb foo.py > foo2.py

In place edit:

ruby -i copyright.rb foo.py