/usr/share/link-grammar/he/4.0.regex is in link-grammar-dictionaries-all 5.3.14-1.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 | %***************************************************************************%
% %
% Copyright (C) 2013 Linas Vepstas %
% See file "LICENSE" for information about commercial use of this system %
% %
%***************************************************************************%
% Want to match apostrophes, for abreviations (I'm I've, etc.) since these
% cannot be auto-split with the current splitter. Also want to accept
% hyphenated words, and word with underbars in them.
%#ANY-WORD: /^[[:alnum:]_'-]+$/
%#ANY-PUNCT: /^[[:punct:]]+$/
% Match anything that doesn't match the above.
% Match anything that isn't white-space.
% Well ... actually, reject anything that begins or ends with
% punctuation. We do this, so that tokenize can split off the
% the affixes (trailing commas, etc) correctly.
%#JUNK: /^[^[:punct:]][^[:space:]]+[^[:punct:]]$/
|