Regular Expressions in Ruby

• published on January 17th, 2008

  PHP implements both POSIX and Perl-compatible regular expressions. The Perl-compatible regexp functions (which includes all the preg_* functions) is the preferred library for most developers since it has many features not available in POSIX, and is binary safe.

  Ruby uses Perl-compatible regular expressions, so if you’re familiar with the preg_* functions in PHP, you’re already well on your way to learning regular expressions in Ruby. Regular expressions are a complex topic, so we won’t be covering regular expression basics, but will instead focus on translating existing knowledge of Perl-compatible PHP functions to Ruby.

  Regular Expressions in Ruby

  We use regular expression patterns in PHP by passing a string argument to various functions. Ruby treats regular expressions differently. Instead of specifying the pattern within a string, they are objects just like everything else in Ruby.

  PHP

  $myRegexp = '/[a-z0-9]+\s/mi';
  print gettype($myRegexp); 
  // => string

  Ruby

  my_regexp = /[a-z0-9]+\s/mi
  p my_regexp.class
  # => Regexp

  We can create regular expressions in Ruby using two different literal syntaxes.
  The most common is by enclosing the pattern in forward-slashes, but we can also use an alternate %r{} syntax. We usually use %r{} when the pattern contains a lot of forward-slashes (such as a filepath). Regular expressions can also be explicitly instantiated using the Regexp class.

  Ruby

  /[a-z0-9]+\s/mi
  %r{/path/to/gif\.gif}mi
  Regexp.new("[a-z0-9]+\s", Regexp::IGNORECASE | Regexp::MULTILINE)

  Comparing Functions/Methods

  Now that we’ve seen some basic syntax for regular expression objects in Ruby, let’s take a look at PHP’s PCRE functions, and their closest equivalents in Ruby.

  PHP	Ruby
  preg_match	String#match
  preg_match_all	String#scan
  preg_replace	String#gsub
  preg_split	String#split
  preg_grep	Array#grep
  preg_quote	Regexp.escape

  preg_match vs. String#match

  We match a pattern in Ruby strings using the match method. Ruby’s match method works differently than preg_match in how it returns matches. We usually want to know two different things when we match data: If the pattern matched, and what specific strings sections were matched.

  PHP returns an integer to tell us if the data matched (either 0 or 1) and populates a matches array by reference. Ruby returns a MatchData object when the pattern matches, and nil when something doesn’t. We can inspect the MatchData object to find the actual string matches.

  In this example, we try to match the different components of a list of email addresses. Both preg_match and String#match only match the first occurrence of the pattern.

  PHP

  $string = 'joe@example.com; walter@example.org';
  $result = preg_match('/([a-z0-9_.-]+)@([a-z0-9-]+)\.([a-z.]+)/i', 
            $string, $matches);
  var_export($result); 
  // => 1
   
  var_export($matches); 
  // => array('joe@example.com', 'joe', 'example', 'com')

  Ruby

  string = 'joe@example.com; walter@example.org'
  matches = string.match(/([a-z0-9_.-]+)@([a-z0-9-]+)\.([a-z.]+)/i)
  p !matches.nil?
  # => true
   
  p matches
  # => #<MatchData:0x1ed138>
   
  p matches[1]
  # => "joe"
   
  p matches.to_a
  # ["joe@example.com", "joe", "example", "com"]

  preg_match_all vs. String#scan

  PHP returns an integer with the number of matches for preg_match_all and populates a matches array by reference. Ruby performs multiple matches for a string using the scan method. This method returns a nested array of matches or an empty array when no matches are found. Be aware that the nesting of values in this array is different than how preg_match_all orders matches.

  In this example, we match components of the email address, and both preg_match_all and String#scan give us an array of matches that are found.

  PHP

  $string = 'joe@example.com; walter@example.org';
  $result = preg_match_all('/([a-z0-9_.-]+)@([a-z0-9-]+)\.[a-z.]+/i', 
            $string, $matches);
  var_export($result);
  // => 2
   
  var_export($matches);
  // => array(array('joe@example.com', 'walter@example.org'),
  //          array('joe',             'walter'),
  //          array('example',         'example'),

  Ruby

  string = 'joe@example.com; walter@example.org'
  result = string.scan(/([a-z0-9_.-]+)@([a-z0-9-]+)\.[a-z.]+/i)
  p result.size
  # => 2
   
  p result
  # => [["joe", "example"], ["walter", "example"]]

  preg_replace vs. String#gsub

  We perform pattern based substitution in Ruby using gsub, which is equivalent to PHP’s preg_replace function. A notable difference is that the gsub method is also used for string substitution in Ruby. We do this by simply providing a string instead of the regular expression pattern. This would be like using the str_replace function in PHP.

  In this example, we want to replace the domain in all emails with foo.

  PHP

  $string = 'joe@example.com; walter@example.org';
  $result = preg_replace('/@([a-z0-9-]+)/', '@foo', $string);
   
  var_export($result);
  // => 'joe@foo.com; walter@foo.org'

  Ruby

  string = 'joe@example.com; walter@example.org'
  result = string.gsub(/@([a-z0-9-]+)/, '@foo')
   
  p result
  # => "joe@foo.com; walter@foo.org"

  We can use backreferences in our gsub replacements just as we would with preg_replace by using \1, \2, etc in our replacement string.

  In this example, we prefix the existing domain with mail.. Remember to escape backslashes used for the backreference.

  PHP

  // Replace domain with mail.domain
  $string = 'joe@example.com; walter@example.org';
  $result = preg_replace('/@([a-z0-9-]+)/', '@mail.\\1', $string);
   
  var_export($result);
  // => 'joe@mail.example.com; walter@mail.example.org'

  Ruby

  string = 'joe@example.com; walter@example.org'
  result = string.gsub(/@([a-z0-9-]+)/, '@mail.\\1')
   
  p result
  # => "joe@mail.example.com; walter@mail.example.org"

  preg_split vs. String#split

  We split strings by a pattern in Ruby using the split method. This is pretty much the same as the preg_split function in PHP. As with gsub, we can also use this same method to split using a string instead of a regular expression. This means that split also performs the equivalent of the explode function in PHP.

  In this example, we create an array of the list of emails by splitting the string using the semi-colon and space as the delimiter.

  PHP

    $string = 'joe@example.com; walter@example.org';
    $result = preg_split('/;\s?/', $string);
   
    var_export($result);
    // array('joe@example.com', 'walter@example.org')

  Ruby

    string = 'joe@example.com; walter@example.org'
    result = string.split(/;\s?/)
   
    p result
    # => ["joe@example.com", "walter@example.org"]

  preg_grep vs. Array#grep

  The preg_grep function in PHP is a useful function to find entries in an array that match a given pattern. Ruby does this same operation with the grep method.

  In this example, we’ll build a new array that only consists of email addresses that end in .com.

  PHP

  $myArray = array('joe@example.com', 'walter@example.org');
  $result = preg_grep('/\.com$/', $myArray);
   
  var_export($result);
  // => array('joe@example.com')

  Ruby

  my_array = ['joe@example.com', 'walter@example.org']
  result = my_array.grep(/\.com$/)
   
  p result
  # => ["joe@example.com"]

  preg_quote vs. Regexp.quote

  When we use a string as a regular expression, we want to escape the characters that could be interpreted as regexp special characters. PHP does this using preg_quote, and Ruby has an equivalent Regexp.escape method.

  In this example, we’ll escape any regular expression special character in the given string.

  PHP

  $string = '[my_file.gif]';
  $result = preg_quote($string);
   
  var_export($result);
  // => '\\[my_file\\.gif\\]'

  Ruby

  string = '[my_file.gif]'
  result = Regexp.escape(string)
   
  p result
  # => "\\[my_file\\.gif\\]"

  Regular Expressions in Rails

  Rails uses regular expressions in various places to specify patterns. When we are matching a route in Rails, we can use them to assign a requirement that a route component must match:

  Ruby

  ActionController::Routing::Routes.draw do |map|
    map.connect 'teams/:team_id/players/:action/:id, :team_id => /\d+/
  end

  We can also use regular expressions in our models when we validate the format of data. We pass a regexp to the :with option of validates_format_of:

  Ruby

  class Image < ActiveRecord::Base
    validates_format_of :url, :with => /\.(gif|jpg)/i, 
                        :message => "must be a GIF or JPG" 
  end

  When we are testing controller code, the assert_select method will accept a regular expression to match response data according to the given pattern.

  Ruby

  class HomepageControllerTest < ActionController::TestCase
    def test_greeting
      get :index
      assert_select 'div.greeting', /Welcome [a-z0-9-_]+/
    end
  end

This entry was posted on Thursday, January 17th, 2008 at 6:38 pm. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

Subscribe

Recent Posts

Get the Book

Talk to Us

Our Company