Rooba
/
spaCy
mirror of https://github.com/explosion/spaCy.git
1
1
Fork 0
Code Issues Packages Projects Releases Wiki Activity

* Add some ad hoc regexes, for multi-word location prepositions

This commit is contained in:
Matthew Honnibal 2015-04-17 04:44:24 +02:00
parent f7ffd94e6a
commit cc4e395927
1 changed files with 7 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
spacy/en/regexes.py
View File

@ -17,6 +17,7 @@ _mw_prepositions = [
'on our way', 'on our way',
'on their way to', 'on their way to',
'on their way', 'on their way',
'along the route from'
] ]
@ -29,6 +30,11 @@ TIME_RE = re.compile(
one_two_digits=r'[0-2]?[0-9]', one_two_digits=r'[0-2]?[0-9]',
am_pm=r'[ap]\.?m\.?')) am_pm=r'[ap]\.?m\.?'))
DATE_RE = re.compile(
'(?:this|last|next|the) (?:week|weekend|{days})'.format(
days='Monday|Tuesday|Wednesday|Thursday|Friday|Saturday|Sunday'
))
MONEY_RE = re.compile('\$\d+(?:\.\d+)?|\d+ dollars(?: \d+ cents)?') MONEY_RE = re.compile('\$\d+(?:\.\d+)?|\d+ dollars(?: \d+ cents)?')
@ -37,4 +43,5 @@ DAYS_RE = re.compile('Monday|Tuesday|Wednesday|Thursday|Friday|Saturday|Sunday')
REGEXES = [('IN', 'O', MW_PREPOSITIONS_RE), ('CD', 'TIME', TIME_RE), REGEXES = [('IN', 'O', MW_PREPOSITIONS_RE), ('CD', 'TIME', TIME_RE),
('NNP', 'DATE', DATE_RE),
('NNP', 'DATE', DAYS_RE), ('CD', 'MONEY', MONEY_RE)] ('NNP', 'DATE', DAYS_RE), ('CD', 'MONEY', MONEY_RE)]