From 04c4064c68a04954c541bf122bcd913fd23ff2da Mon Sep 17 00:00:00 2001
From: 03 <1930154319@qq.com>
Date: Mon, 18 Mar 2024 13:21:46 +0800
Subject: [PATCH] Fix bug 553
---
app/templates/mainpage_post.html | 2 +-
app/wordfreqCMD.py | 4 +++-
2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/app/templates/mainpage_post.html b/app/templates/mainpage_post.html
index 7357457..5df7dd5 100644
--- a/app/templates/mainpage_post.html
+++ b/app/templates/mainpage_post.html
@@ -2,7 +2,7 @@
- Title
+ 单词词频
{{ yml['header'] | safe }}
{% if yml['css']['item'] %}
diff --git a/app/wordfreqCMD.py b/app/wordfreqCMD.py
index dcee74e..dfdde2e 100644
--- a/app/wordfreqCMD.py
+++ b/app/wordfreqCMD.py
@@ -4,6 +4,7 @@
###########################################################################
import collections
+import html
import string
import operator
import os, sys # 引入模块sys，因为我要用里面的sys.argv列表中的信息来读取命令行参数。
@@ -39,7 +40,8 @@ def file2str(fname):#文件转字符
def remove_punctuation(s): # 这里是s是形参 (parameter)。函数被调用时才给s赋值。
- special_characters = '\_©~<=>+/[]*&$%^@.,?!:;#()"“”—‘’{}|' # 把里面的字符都去掉
+ special_characters = '\_©~<=>+/[]*&$%^@.,?!:;#()"“”—‘’{}|，。？！￥……（）、《》：；·' # 把里面的字符都去掉
+ s = html.unescape(s)
for c in special_characters:
s = s.replace(c, ' ') # 防止出现把 apple,apple 移掉逗号后变成 appleapple 情况
s = s.replace('--', ' ')