fix cellphone extraction bugs

Author: Yang Yang

cocoNLP/__version__.py

@@ -10,7 +10,7 @@
 __title__ = "cocoNLP"
 __description__ = "Python implementation of many nlp algorithms"
 __url__ = "https://github.com/fighting41love"
-__version__ = "0.0.11"
+__version__ = "0.0.12"
 __author__ = "Yang Yang"
 __author_email__ = "yangyangfuture@gmail.com"
 __license__ = "MIT"

cocoNLP/extractor.py

@@ -61,11 +61,11 @@ def extract_cellphone(self, text, nation):
         :return: email_addresses_list<list>
         """
         eng_texts = self.replace_chinese(text)
-        sep = ',!?:; ：，。！？《》、|\\/abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
+        sep = ',!?:; ：，.。！？《》、|\\/abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
         eng_split_texts = [''.join(g) for k, g in groupby(eng_texts, sep.__contains__) if not k]
         eng_split_texts_clean = [ele for ele in eng_split_texts if len(ele)>=7 and len(ele)<17]
         if nation=='CHN':
-            phone_pattern = '((\+86)?([- ])?)?(|(13[0-9])|(14[0-9])|(15[0-9])|(17[0-9])|(18[0-9])|(19[0-9]))([- ])?\d{3}([- ])?\d{4}([- ])?\d{4}'
+            phone_pattern = r'^((\+86)?([- ])?)?(|(13[0-9])|(14[0-9])|(15[0-9])|(17[0-9])|(18[0-9])|(19[0-9]))([- ])?\d{3}([- ])?\d{4}([- ])?\d{4}$'
 
         phones = []
         for eng_text in eng_split_texts_clean:

test.py

@@ -3,7 +3,8 @@
 
 ex = extractor()
 
-text = '急寻特朗普，男孩，于2018年11月27号11时在陕西省安康市汉滨区走失。身份证号码410105196904010537丢失发型短发，...如有线索，请迅速与警方联系：18100065143，132-6156-2938，baizhantang@sina.com.cn 和yangyangfuture at gmail dot com'
+text = '急寻特朗普，男孩，于2018年11月27号11时在陕西省安康市汉滨区走失。身份证号码410105196904010537丢失发型短发，...' \
+       '如有线索，请迅速与警方联系：18100065143，132-6156-2938，baizhantang@sina.com.cn 和yangyangfuture at gmail dot com13673630861'
 
 # 抽取邮箱
 emails = ex.extract_email(text)