Matching Regular Expressions

Perl regular expressions look very much like those in vi.

.

Match any one character except a newline.

c*

Match zero or more instances of character c.

c+

Match one or more instances of character c.

c?

Match zero or one instance of character c.

[class]

Match any of the characters in character class class.

\w

Match an alphanumeric character (including "_")

\W

Match an non-alphanumeric character (including "_")

\b

Match a word boundary

\B

Match non-boundaries

\s

Match a whitespace character

\S

Match a non-whitespace character

\d

Match a numeric character

\D

Match a non-numeric character

^ Match the beginning of a line

$

Match the end of a line

Also, \n, \r, \f, \t and \NNN have their usual C-style interpretations.

The actual syntax for the pattern matching command is m/pattern/gio. The modifiers are g for ``global'' match, i for ``ignore case'', and o for ``only compile this regexp once''. With the m command, you can use any pair of non-alphanumeric characters to bound the expression. This especially useful when matching filenames that contain the ``/'' character. For example:

if (m!^/tmp_mnt!)
{ print "$_ is an automounted file system\n"; }

If the delimiter you choose is ``/'', then the leading m is optional.

Perl even has the ability to do multi-line pattern matching. Refer to the documentation on the $* variable for complete information.