From 1faaf698ca3fc33eb4bf8fc9e8ae87d4ec582486 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Rapha=C3=ABl=20Bournhonesque?= <raphael@likeabird.io>
Date: Tue, 24 Jan 2017 09:51:29 +0100
Subject: [PATCH] Add infixes and abbreviation exceptions (fr)

---
 spacy/fr/language_data.py | 51 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 51 insertions(+)

diff --git a/spacy/fr/language_data.py b/spacy/fr/language_data.py
index bbbeb1535..6eff8364f 100644
--- a/spacy/fr/language_data.py
+++ b/spacy/fr/language_data.py
@@ -4,6 +4,9 @@ from __future__ import unicode_literals
 from .. import language_data as base
 from ..language_data import strings_to_exc, update_exc
 
+from .punctuation import ELISION
+
+from ..symbols import *
 from .stop_words import STOP_WORDS
 
 
@@ -13,5 +16,53 @@ STOP_WORDS = set(STOP_WORDS)
 TOKENIZER_EXCEPTIONS = strings_to_exc(base.EMOTICONS)
 update_exc(TOKENIZER_EXCEPTIONS, strings_to_exc(base.ABBREVIATIONS))
 
+ABBREVIATIONS = {
+    "janv.": [
+        {LEMMA: "janvier", ORTH: "janv."}
+    ],
+    "févr.": [
+        {LEMMA: "février", ORTH: "févr."}
+    ],
+    "avr.": [
+        {LEMMA: "avril", ORTH: "avr."}
+    ],
+    "juill.": [
+        {LEMMA: "juillet", ORTH: "juill."}
+    ],
+    "sept.": [
+        {LEMMA: "septembre", ORTH: "sept."}
+    ],
+    "oct.": [
+        {LEMMA: "octobre", ORTH: "oct."}
+    ],
+    "nov.": [
+        {LEMMA: "novembre", ORTH: "nov."}
+    ],
+    "déc.": [
+        {LEMMA: "décembre", ORTH: "déc."}
+    ],
+}
+
+
+INFIXES_EXCEPTIONS_BASE = ["aujourd'hui",
+                           "prud'homme", "prud'hommes",
+                           "prud'homal", "prud'homaux", "prud'homale",
+                           "prud'homales",
+                           "prud'hommal", "prud'hommaux", "prud'hommale",
+                           "prud'hommales",
+                           "prud'homie", "prud'homies",
+                           "prud'hommesque", "prud'hommesques",
+                           "prud'hommesquement"]
+
+INFIXES_EXCEPTIONS = []
+for elision_char in ELISION:
+    INFIXES_EXCEPTIONS += [infix.replace("'", elision_char)
+                           for infix in INFIXES_EXCEPTIONS_BASE]
+
+INFIXES_EXCEPTIONS += [word.capitalize() for word in INFIXES_EXCEPTIONS]
+
+update_exc(TOKENIZER_EXCEPTIONS, strings_to_exc(INFIXES_EXCEPTIONS))
+update_exc(TOKENIZER_EXCEPTIONS, ABBREVIATIONS)
+
 
 __all__ = ["TOKENIZER_EXCEPTIONS", "STOP_WORDS"]