6.2 Word lists

• word lists can be stored in variables

  > define $week =
"Monday Tuesday Wednesday Thursday Friday";

  and used instead of regular expressions in the attribute/value pairs

  > [lemma = $week];

  (word lists are not allowed in XML start tags, though)

• add/delete words with += and -=

  > define $week += "Saturday Sunday";

• show list of words stored in variable

  > show $week;

  use show var; to see all variables

• read word list from file (one-word-per-line format)

  > define $week < "/home/weekdays.txt";

  new in CQP v3.4.11: files ending in .gz or .bz2 are automatically decompressed (see Sec. 3.1), and word lists can be read from a shell pipe indicated by a | character at the start of the filename

  for example, to read a file with whitespace-delimited words (and multiple entries per line):

  > define $week < "| perl -pe 's/\s+/\n/g' words.txt";

• use TAB key to complete variable names (e.g. type “show $we” + TAB)

• word lists can be used to simulate type hierarchies, e.g. for part-of-speech tags

  > define $common_noun = "NN NNS";
> define $proper_noun = "NP NPS";

  > define $noun = $common_noun;
> define $noun += $proper_noun;

• %c and %d flags can not be used with word lists

• use lists of regular expressions with the RE() operator (compile regex)

  > define $pref="under.+ over.+";
> [(lemma=RE($pref)) & (pos="VBG")];

• flags can be appended to the RE() operator

  > [word = RE($pref) %cd];