2.2 Searching for words

• search single word form (single or double quotes are required: '...' or "...")

  > "interesting";

  $\to$ shows all occurrences of interesting

• the specified word is interpreted as a regular expression

  > "interest(s|(ed|ing)(ly)?)?";

  $\to$ interest, interests, interested, interesting, interestedly, interestingly

• see Appendix A.1 for an introduction to the regular expression syntax

• the regular expression “flavour” used by CQP is Perl Compatible Regular Expressions, usually known as PCRE; lots of documentation and examples can be found on the WWW

• note that special characters have to be “escaped” with backslash (\)

  "?" fails;$\quad$ "\?" $\to$ ?;$\quad$ "." $\to$ . , ! ? a b c ...;$\quad$ "\$\." $\to$ $.

  “critical” characters are: . ? * + | ( ) [ ] { } ^ $

• CWB 3.0: L^ATEX-style escape sequences \", \', \` and \^, followed by an appropriate ASCII letter, are used to represent characters with diacritics when they cannot be entered directly

  "B\"ar" $\to$ Bär;$\quad$ "d\'ej\`a" $\to$ déjà

  NB: this feature is deprecated; it works only for the Latin-1 encoding and cannot be deactivated

• CWB 3.0: additional special escape sequences:

  \"s $\to$ ß;$\quad$ \,c $\to$ ç;$\quad$\,C $\to$ Ç;$\quad$ \~n $\to$ ñ;$\quad$\~N $\to$ Ñ;$\quad$

  version 3.0.3 introduces two-digit hex escapes for inserting arbitrary byte values:

  \xDF $\to$ ß in a Latin1-encoded corpus;$\quad$ \xC3\x9F $\to$ ß in a UTF-8-encoded corpus

• CWB 3.5: full support for PCRE regular expressions, including two- and four-digit hex escapes

• use flags %c and %d after a regex to ignore case / diacritics

  DICKENS> "interesting" %c;
GERMAN-LAW> "wahrung" %cd;

• if you need to match a word form containing single or double quotes (e.g. 'em or 12”-screen), there are two possibilities:
  • if the string does not contain both single and double quotes, simply pick an appropriate quote character: "'em" vs. '12"-screen'
  • otherwise, escape the quote characters with a backslash: '\'em' and "12\"-screen"
  • alternatively, double every occurrence of the quote character inside the string; our two examples could also be matched with the queries ”'em' and "12""-screen"