Tag which defines a group of patterns. It contains the tag patterns, which associate individual definitions of patterns within the same group. See the example for more details.

Dividing parsing rules into groups allows for the next level of rule prioritization. The parsing engine takes each group in turn and takes all the rules from it and tries to match them against the input string. The parsing ends the moment when any rule is found or no more groups exist. If a group is found, it is processed without any relation to any other already processed group nor to any other group still to be processed.

Structuring rules into isolated groups might thus lead to situations where more possible rules found in several groups with only one matching rule in the first matching group result in not setting the GP_MORE_PATTERNS flag.