Quoting Expressions
A string in "" expends special characters
A string in '' is as-is
use q for single quote
q|this is to be quoted|
use qq for double quote
qq~this is to be double quoted~
Pattern Matching
if ($_ =~ /pattern/), which is the same as if (/pattern/)
Not find pattern
(!/pattern/)
Substitution
$_ =~ s/pattern/newpattern/g
Case insensitive
if (/pattern/i)
A set of characters
[ABCDE], [A-Z], or [A-Z0-9]
Not a set of characters
[^ABC]
One or the other
$s =~ /^(A|B)$/
Special shortcuts
| \d == [0-9] | \D == non-digit |
| \w == [0-9A-Za-z] | \W == non-word |
| \s == [\t\n\r] | \S == non-white space |
| \h == horizontal white space | |
| \v == vertical white space | |
| \R == line ending | |
| \N == non-new line | |
| . == any character | |
| \b == word boundary |
This or That
th(is|at)
Repeat Qualifiers after a character
| ? | - zero or 1 |
| * | - zero or more |
| + | - one or more |
| {x,y} | - between x and y times |
| {x} | - exactly x times |
| *? | - zero or more but non-greedy |
| +? | - one or more but non-greedy |
Others
(abc) the string "abc"
Extract matched string
Use of parenthesis
(pattern1)(pattern2)(pattern3) will be extracted to $1, $2, and $3
use the trailing /x in a regular expression to divid the expression in separate lines
use the trailing /o in a regular expression to ask Perl not to recompile
Parameter Matching and Capture
Parameters are captured by enclosing a matching expression in parenthesis. That is, (\S*)...(\w+). The parameters can then be accessed using $1, $2, $3, etc..
Ex: $d =~ s/\d{2}(\d{2})(\d{2})(\d{2})/$2\/$3\/$1/; #date translation
Add label to expression
(\S+) becomes (?<label>\S+) and can be referenced by $+{label} - a hash
To allow a non-capture parenthesis
(?:....)
Use non-greedy match to avoid performance penalty
Avoid using the match variables $`, $&, and $', or use the /p flag
Once these match variables are used, they incur a performance penalty on subsequent matches. To specify that these variables should only be set for the current match, use the suffix /p int eh expression - m/..../p.
Performance
To test performance of regexp, use benchmark
use Benchmark qw(timethese);
timethese {
........
}
