RegEx Introduction and Quick Reference

Note: Most of this info is available directly from the Zend Website, there are a couple of original additions but most of it is ripped with comments.

A list of all the PCRE (Perl-Compatible Regular Expression) functions are available with full documentation in the PHP Manual, here's a quick link to the pertinent page.

Basic Examples using Metacharacters

• preg_match( '/apple/', 'I saw the apple tree.' );

  • Returns: True.

• preg_match( '/a..le/', 'I saw the apple tree' );

  • Returns: True.
  • Note: . = any char except new line.

• preg_match( '/^apple/', 'I saw the apple tree.' );

  • Returns: False.
  • Note: ^ = start of string.

• preg_match( '/apple$/', 'I saw the apple tree.' );

  • Returns: False.
  • Note: $ = end of string.

• preg_match( '/appleX?/', 'I saw the apple tree.' );

  • Returns: True.
  • Note: ? = prior bit can occur 0 or 1 times, X? == (X)?, XX? != (XX)?

• preg_match( '/ap*le/', 'I saw the apple tree.' );

  • Returns: True.
  • Note: * = 0 or more of prior bit (so appppple would match).

• preg_match( '/ap+le/', 'I saw the apple tree.' );

  • Returns: True.
  • Note: + = like * but 1 or more of the prior bit.

• preg_match( '/ap{2}le/', 'I saw the apple tree.' );

  • Returns: True.
  • Note: {x} = prior bit MUST match x times.

• preg_match( '/ap{2,4}le/', 'I saw the apple tree.' );

  • Returns: True.
  • Note: {x,z} = prior bit can match any amount of times between x & z inclusive (so 2,3 or 4 in the eg).

• preg_match( '/(apple|orange)/', 'I saw the apple tree.' );

  • Returns: True.
  • Note: | = OR

• preg_match( '/apple[0-9m-z]/', 'I saw the apples tree.' );

  • Returns: True.
    
  • Note: [] = range of chars. [0-9a-zA-Z] would match every char/numeral.

• preg_match( '/ap*?le/', 'I saw the apple tree.' );

  • Returns: True.
  • Note: *? = ungreedily match 0 or more of prior bit (so appppple would match), but when returning matches the least possible number of characters would be returned. See the ungreedy modifier below.

Extra note on the use of ? after metacharacters that specify matching a number of the prior bit, ie. ?, *, +, {}.

Special (Meta) Characters

• \s = used to match a single white space character, including tabs and newline characters.
• \S = used to match everything that is *not* a white space character.
• \b = used to match word boundaries
• \B = used to match anything that does not match with \b
• \d = used to match numbers from 0 to 9.
• \D = used to match anything that does not match with \d
• \w = used to match letters, numbers and underscores.
• \W = used to match anything that does not match with \w
• \n = newline
• \r = carriage return

Note

Escaping also needs to be used when using any of the meta characters mentioned above, you can't just match '1 + 2' with '/1 + 2/', you have to escape the '+'. So use '1 \+ 2'. Remember that as '\' denotes an escape sequence, it too needs escaping '\\'.

Modifiers

• i = case Insensitive
  • Does exactly what it says on the tin.
• U = Ungreedy
  • A hard one to explain; take BB code tags for example - it helps because it won't screw up '[b]bold1[/b] blah [b]bold2[/b]' by not being greedy enough to take the first [b] and the last [/b], instead it parses it as you'd expect. Note that it's better to use the specific case with this, ie. doing '#\[b\](.+?)\[/b\]#i' Note that regex searches are greedy by default, and using this modifier reverses that.
• s = multiline mode
  • This makes a difference to whether ^ and $ refer to things on the CURRENT LINE or the WHOLE string (ie. when 's' is used $ = end of string, not end of line)

Submatching

• preg* is faster than ereg*, so use those functions. Ereg* may appear at first to be simpler, but in my opinion they're just not as easy in the long run.

• $string = 'how now, brown cow?';
  $string = preg_replace( '/^(.+)\snow, (.+)$/', 'first submatch: "\\1", second submatch: "\\2"', $string );
  echo $string;

  • Returns: 'first submatch: "how", second submatch: "brown cow?"'
  • Note: Submatching is tremendously useful for any medium/advanced level projects. ie. Most projects that would require regexs.
  • Note: I only used '\s' as an example of white space, note that I used a normal space on the other side of 'now,'. Pay close attention to the fact that the string 'snow' DOES NOT OCCUR inside $string.

• $string = 'how now, brown cow?';
  preg_match( '/^(.+)\snow, (.+)$/', $string, $matches );
  print_r( $matches );

  • Returns:

    Array
    (
        [0] => how now, brown cow?
        [1] => how
        [2] => brown cow?
    )

  • As you can see, the entirety of the string subsection looked at by the regex is in array index 0.
  • Use of print_r() is of course an exceptionally useful method of debugging your regexs.

Notes