Grouping and Captured Groups

Grouping alternatives

A better syntax for \bJane\b|\bJanet\b is \b(Jane|Janet)\b using () for grouping alternatives. However this also creates a captured group, if you don't need any captured group to reuse, see Noncapturing

Noncapturing groups

\b(?:Jane|Janet)\b disables group capturing using (?:), it won't capture the group when matching, benefit to better performance.

Group with mode modifier

Use any of (?i:<regex>) or (?s:<regex>) or (?m:<regex>) to annotate grouped alternatives

• \b(?i:Jane|Janet)\b

To combine different modes:

• (?ism:<regex>) enables case-insensitive, singleline and multiline
• (?-ism:<regex>) disables case-insensitive, singleline and multiline
• (?i-sm:<regex>) enables case-sensitive and disables singleline and multiline

Backreference

Reference the first nine capturing groups with a backslash followed by a single digit one through nine(\1 to \9). For groups 10 through 99, use \10 to \99 The following regex matches date like 2008-08-08 and 2007-07-07.

\b\d\d(\d\d)-\1-\1\b

If backreferences are before any capturing groups, they fails to match anything.

Named group and named backrefernces

Assign a descriptive name for groups instead of numbers.

The previous example can be written as following:

\b\d\d(?<tailing>\d\d)-\k<tailing>-\1\b

Named group flavors:

flavor	syntax
.NET, Java 7, PCRE 7, Perl 5.10, Ruby 1.9	(?<name>regex)
.NET, PCRE 7, Perl 5.10, Ruby 1.9	(?'name'regex)
Python, PCRE 4, Perl 5.10, RUby 1.9	(?P<name>regex)

Backreference flavours:

flavor	syntax
.NET, Java 7, PCRE 7, Perl 5.10, Ruby 1.9	\k<name>
.NET, PCRE 7, Perl 5.10, Ruby 1.9	\k'name'
Python, PCRE 4, Perl 5.10, Ruby 1.9	(?P=name)

INFO

<> are not placeholder in tables above.