Textual regular expressions

In CQL, regular expressions are most commonly used in the ⊢ filter to search for sequences of positions. In the non-CQL world, regular expressions are most commonly used to search text for sequences of characters. See regular expressions . We call these regular expressions designed to search in text textual regular expressions to distinguish them from the chess kind.

A texual regular expression is just a sequence of alphanumeric characters and spaces. Each textual regular expression will either match or not match a given string.

Some more examples: Bor?gov would match either Borgov or Bogov, but not Borrgov, since a ? denotes 0 or one repetition

Bor*gov would match Bogov or Borgov or Borrgov or Borrrgov and so on, because * denotes 0 or more repetitions.

Likewise, Bor+gov would match Borrrrgov but not Bogov as at least one repetition is needed.

Bo[r-t]gov would match Borgov or Bosgov or Botgov, just like in piece designators.

Bor{3,8}gov would match Borrrrrgov, anything with betweeen 3 and 8 rs.

Bo.gov would match Boxgov or Bo_gov or Bo gov . Here, . matches any character (just like in the rest of CQL, the . matches any square or any position).

Borgov[0-9]+ would match Borgov followed by a nonempty sequence of digits, like Borgov0358.

The parentheses denotes grouping, e.g. Bor(go)+v would match Borgov and Borgogov and Borgogogov and so on.

Escaping special characters in regular expressions

Some characters have special meanings within textual regular expressions, namely

	  {}()[].?+*\^$

To use such a character literally inside a regular expression, precede it with a backslash. For example, to match the three character string (c), use the regular expression \(c\). To search for a literal period in an an event, use \. instead of . .

other special characters

There are host of more obscure regular expression facilities, like \w, \W and many others. The two most common such facilities useful in CQL are \d, which matches any digit (is thus equivalent to [0-9] ); and \n , which matches a newline character.

version of regular expressions used

CQL currently uses the default ECMAScript variant of regular expressions as interpreted in C++. However, we recommend that users generally use only the basic regex syntaxes discussed here unless they are already familiar with this variant. Moreover, we have observed some lack of uniformity in the implementation of this standard among the different C++ compilers on which CQL is typically compiled.

Comparison of textual regular expressions to regular expression with ⊢

The meanings of textual regular expression operators ∙, * in ⊢, {} operator, + in ⊢, parentheses in ⊢, ? in ⊢ are the same as in ⊢, except for matching characters rather than positions. Note that the regular expression . is now the center dot ∙ for compatibility reasons.

Brackets are used in piece designators similar to how they are in textual regular expressions.

The | textual regular expression operator corresponds to the CQL or filter and is not a regular expression operator in ⊢.