update rules for clean_text

ZJ-CAI · Aug 18, 2021 · 42761e2 · 42761e2
1 parent e1742cc
commit 42761e2
Show file tree

Hide file tree

Showing 6 changed files with 10 additions and 9 deletions.
diff --git a/README.md b/README.md
@@ -11,7 +11,7 @@
     <a alt="Downloads">
         <img src="https://img.shields.io/badge/downloads-4k-yellow" /></a>
     <a alt="Version">
-        <img src="https://img.shields.io/badge/version-1.3.27-green" /></a>
+        <img src="https://img.shields.io/badge/version-1.3.28-green" /></a>
     <a href="https://github.com/dongrixinyu/JioNLP/pulse" alt="Activity">
         <img src="https://img.shields.io/github/commit-activity/m/dongrixinyu/JioNLP?color=blue" /></a>
 </p>
@@ -95,6 +95,7 @@ $ pip install .
 $ pip install jionlp
 ```
 
+
 ## 使用 Features
 
 - 导入工具包，查看工具包的主要功能与函数注释
@@ -230,7 +231,7 @@ $ jio_help
 - 如感兴趣合作完善本工具包，请参考 **TODO.txt** 文件进行功能添加。
 
 ### 做 NLP不易，欢迎加入自然语言处理 Wechat 交流群
-### 请先添加vx：dongrixinyu89
+### 如以下码失效，请先添加vx：dongrixinyu89
 ![image](../../blob/master/image/qr_code_for_collection.png)
 
 
diff --git a/image/qr_code_for_collection.png b/image/qr_code_for_collection.png
diff --git a/jionlp/__init__.py b/jionlp/__init__.py
@@ -8,7 +8,7 @@
 # description: Preprocessing tool for Chinese NLP
 """
 
-__version__ = '1.3.27'
+__version__ = '1.3.28'
 
 import os
 

diff --git a/jionlp/gadget/time_parser.py b/jionlp/gadget/time_parser.py
@@ -4473,14 +4473,14 @@ def string2handler(datetime_obj):
                     29, leap_month)  # 当农历月无30 天时，按 29天计算
 
             return string2handler(first_solar_time_handler),\
-                   string2handler(second_solar_time_handler)
+                string2handler(second_solar_time_handler)
 
         else:
             solar_time_handler = self.lunar2solar(
                 lunar_time_handler[0], lunar_time_handler[1],
                 lunar_time_handler[2], leap_month)
             return string2handler(solar_time_handler),\
-                   string2handler(solar_time_handler)
+                string2handler(solar_time_handler)
 
     def _parse_solar_terms(self, year, solar_term):
         """解析24节气

diff --git a/jionlp/rule/rule_pattern.py b/jionlp/rule/rule_pattern.py
@@ -59,7 +59,7 @@
 # 以及相应的常用符号，单字节符号、标点符号等。而日文、俄文、拉丁、希腊、数学公式、
 # 物理单位等符号 绝大多数不常用的都被丢弃。其中 㐀-䶵 指的是另一个汉字字符集
 # 仅保留了常用符号，数字标识，如 ① 等
-UNICODE_EXCEPTION_PATTERN = '[^‐-”•…‰※℃℉Ⅰ-ⅹ①-⒛\u3000-】〔-〞㈠-㈩一-龥﹐-﹫！-～￠￡￥]'
+UNICODE_EXCEPTION_PATTERN = '[^‐-”•·・…‰※℃℉Ⅰ-ⅹ①-⒛\u3000-】〔-〞㈠-㈩一-龥﹐-﹫！-～￠￡￥]'
 EXCEPTION_PATTERN = ASCII_EXCEPTION_PATTERN[:-1] + UNICODE_EXCEPTION_PATTERN[2:]
 
 # ---------------------------------------------------------------------
@@ -155,7 +155,7 @@
 # ---------------------------------------------------------------------
 # 冗余字符处理
 # 文本中有连续的 “哈哈哈哈哈” 等字符串，需要删除冗余字符串，返回为 “哈”
-REDUNDANT_PATTERN = ' -\t\n啊哈呀~\u3000\xa0'
+REDUNDANT_PATTERN = ' -\t\n啊哈呀~\u3000\xa0•·・'
 
 # ---------------------------------------------------------------------
 # 纯数字格式，用于过滤停用词时，过滤掉纯数字（包括汉字数字）
@@ -305,7 +305,7 @@
 # 固定公历节日
 FIXED_SOLAR_FESTIVAL = r'((元旦|十一)|(三八|五一|六一|七一|八一|国庆|圣诞)(节)?|'\
     r'((三八)?妇女|女神|植树|(五一)?劳动|(五四)?青年|(六一)?儿童|(七一)?建党|(八一)?建军|教师|情人|愚人|万圣|护士)节|'\
-    r'地球日|三[\.·]?一五|双(十一|11)|(.{1,4})?消费者权益日)'
+    r'地球日|三[\.•·・]?一五|双(十一|11)|(.{1,4})?消费者权益日)'
 # 固定农历节日
 FIXED_LUNAR_FESTIVAL = r'((春|填仓|上巳|寒食|清明|浴佛|姑姑|财神|下元|寒衣)节|'\
     r'(龙抬头|除夕)|'\

diff --git a/setup.py b/setup.py
@@ -50,7 +50,7 @@
 __maintainer__ = "dongrixinyu"
 __email__ = "[email protected]"
 __url__ = 'https://github.com/dongrixinyu/JioNLP'
-__description__ = 'Preprocessing tool for Chinese NLP'
+__description__ = 'Chinese NLPreprocessing & Parsing'
 
 
 with open(os.path.join(DIR_PATH, 'requirements.txt'),