A..1 Summary of regular expression syntax

At the character level, CQP supports regular expressions using one of two regex libraries:

CWB 3.0: Uses POSIX 1003.2 regular expressions (as provided by the system libraries). A full description of the regular expression syntax can be found on the regex(7) manpage.

CWB 3.5: Uses PCRE (Perl Compatible Regular Expressions). A full description of the regular expression syntax can be found on the pcrepattern(3) manpage; see also http://www.pcre.org/.

Various books such as Mastering Regular Expressions give a gentle introduction to writing regular expressions and provide a lot of additional information. There are also many tutorials to be found online using Your Favourite Web Search Engine^TM.

• A regular expression is a concise descriptions of a set of character strings (which are called words in formal language theory). Note that only certain sets of words with a relatively simple structure can be represented in such a way. Regular expressions are said to match the words they describe. The following examples use the notation:

  <reg.exp.> $\to$ word$_1$, word$_2$, ...

  to indicate that the regular expression before the arrow matches the word or words after the arrow. In many programming languages, it is customary to enclose regular expressions in forward slashes (/). CQP uses a different syntax: regular expressions are written as (single- or double-quoted) strings. The examples below omit any delimiters.

• Basic syntax of regular expressions
  • letters and digits are matched literally (including all non-ASCII characters)
    word $\to$ word;$\quad$C3PO $\to$ C3PO;$\quad$ déjà $\to$ déjà
  • . matches any single character (“matchall”)
    r.ng $\to$ ring, rung, rang, rkng, r3ng, ...
  • character set: [...] matches any of the characters listed
    moderni[sz]e $\to$ modernise, modernize
    [a-c5-9] $\to$ a, b, c, 5, 6, 7, 8, 9
    [^aeiou] $\to$ b, c, d, f, ..., 1, 2, 3, ..., ä, à, á, ...
  • repetition of the preceding element (character or group):
    ? (0 or 1), * (0 or more), + (1 or more), {$n$} (exactly $n$), {$n$,$m$} ($n\ldots m$)
    colou?r $\to$ color, colour;$\quad$ go{2,4}d $\to$ good, goood, goood
    [A-Z][a-z]+ $\to$ “regular” capitalised word such as British
  • grouping with parentheses: (...)
    (bla)+ $\to$ bla, blabla, blablabla, ...
    (school)?bus(es)? $\to$ bus, buses, schoolbus, schoolbuses
  • | separates alternatives (use parentheses to limit scope)
    mouse|mice $\to$ mouse, mice;$\quad$ corp(us|ora) $\to$ corpus, corpora
• Complex regular expressions can be used to model (regular) inflection:
  • ask(s|ed|ing)? $\to$ ask, asks, asked, asking
    (equivalent to the less compact expression ask|asks|asked|asking)
  • sa(y(s|ing)?|id) $\to$ say, says, saying, said
  • [a-z]+i[sz](e[sd]?|ing) $\to$ any form of a verb with -ise or -ize suffix
• Backslash (\) “escapes” special characters, i.e. forces them to match literally
  • \? $\to$ ?;$\quad$\(\) $\to$ ();$\quad$ .{3} $\to$ ...;$\quad$\$\. $\to$ $.
  • \^ and \$ must be escaped although ^ and $ anchors are not useful in CQP