From 2fee67cfa3938576387fe10bd168cc0820a6e6c5 Mon Sep 17 00:00:00 2001
From: Matthew Honnibal <honnibal@gmail.com>
Date: Tue, 7 Apr 2015 03:45:18 +0200
Subject: [PATCH] * Add regular expressions for English multi-word expressions

---
 spacy/en/regexes.py | 35 +++++++++++++++++++++++++++++++++++
 1 file changed, 35 insertions(+)
 create mode 100644 spacy/en/regexes.py

diff --git a/spacy/en/regexes.py b/spacy/en/regexes.py
new file mode 100644
index 000000000..4eb45d87d
--- /dev/null
+++ b/spacy/en/regexes.py
@@ -0,0 +1,35 @@
+import re
+
+
+_mw_prepositions = [
+    'close to',
+    'down by',
+    'on the way to',
+    'on my way',
+    'on his way',
+    'on her way',
+    'on your way',
+    'on our way',
+    'on their way',
+]
+
+
+
+MW_PREPOSITIONS_RE = re.compile('|'.join(_mw_prepositions), flags=re.IGNORECASE)
+
+
+TIME_RE = re.compile(
+    '{colon_digits}|{colon_digits} ?{am_pm}?|{one_two_digits} ?({am_pm})'.format(
+        colon_digits=r'[0-2]?[0-9]:[0-5][0-9](?::[0-5][0-9])?',
+        one_two_digits=r'[0-2]?[0-9]',
+        am_pm=r'[ap]\.?m\.?'))
+
+
+MONEY_RE = re.compile('\$\d+(?:\.\d+)?|\d+ dollars(?: \d+ cents)?')
+
+
+DAYS_RE = re.compile('Monday|Tuesday|Wednesday|Thursday|Friday|Saturday|Sunday')
+
+
+REGEXES = [('IN', 'O', MW_PREPOSITIONS_RE), ('CD', 'TIME', TIME_RE),
+           ('NNP', 'DATE', DAYS_RE), ('CD', 'MONEY', MONEY_RE)]