有备无患：truncatehanzi - 按字数截取中文的 Django template filter

Django 自带的一个模板过滤器（Template Filter）truncatewords，可以按单词数截取字符串生成摘要。不过，truncatewords 只能用空格来分隔单词，对汉字就无能为力了。所以，我写了另一个模板过滤器 truncatehanzi，它通过 unicode 编码来区分西方字符和 CJK 汉字，对包括英语字母在内的西方字符以空格或其他符号来分隔并计算单词数，而对 CJK 汉字则是按字数计算。

使用方法和 truncatewords 一样，

{{ value|truncatehanzi:6 }}

如果 value 为“截取English、Γρεεκ等字母语言和CJK汉字。”，输出“截取English、Γρεεκ等字...”。

代码分为两个部分，截取汉字的脚本 truncate_hanzi.py 和生成过滤器的脚本 filters.py。

truncate_hanzi.py

 # -*- coding: UTF-8 -*-
 
 import re
 from django.utils.encoding import force_unicode
 from django.utils.functional import allow_lazy
 
 def truncate_hanzi(s, num):
     s = force_unicode(s)
     length = int(num)
     if length <= 0:
         return u'...'
     # Set up regex for alphanumeric characters
     # \u00c0-\u02af: Latin
     # \u0370-\u1fff: Greek and alphabet characters in other language 
     re_alnum = re.compile(ur'[a-zA-Z0-9_\-\u00c0-\u02af\u0370-\u1fff]', re.U)
     # Set up regex for hanzi
     # \u3040-\ufaff: CJK characters
     re_hanzi = re.compile(ur'[\u3040-\ufaff]', re.U)
     hanzi = u''
     hanzi_len = 0
     word_temp = u''
     for char in s:
         # Check for alphabet characters
         if re_alnum.match(char):
             word_temp += char
             continue
         if word_temp:
             hanzi += word_temp
             hanzi_len += 1
             word_temp = ''
         # Check for length
         if hanzi_len >= length:
             if not hanzi.endswith('...'):
                 hanzi += '...'
                 break
         # Check for hanzi
         if re_hanzi.match(char):
             hanzi_len += 1
         hanzi += char
     hanzi += word_temp
     return hanzi
 truncate_hanzi = allow_lazy(truncate_hanzi, unicode)
 
 def demo():
     print truncate_hanzi('截取段落工具，支持English、Γρεεκ等字母语言和CJK汉字。', 6)
     print truncate_hanzi('截取段落工具，支持English、Γρεεκ等字母语言和CJK汉字。', 11)
     print truncate_hanzi('截取段落工具，支持English、Γρεεκ等字母语言和CJK汉字。', 20)
 
 if __name__ == '__main__':
     demo()

filters.py

 # -*- coding: UTF-8 -*-
 from django.template import Library
 from django.template.defaultfilters import stringfilter
 
 register = Library()
 
 @stringfilter
 def truncatehanzi(value, arg):
     """
     Truncates a string after a certain number of words including
     alphanumeric and CJK characters.
 
     Argument: Number of words to truncate after.
     """
     from truncate_hanzi import truncate_hanzi
     try:
         length = int(arg)
     except ValueError: # Invalid literal for int().
         return value # Fail silently.
     return truncate_hanzi(value, length)
 truncatehanzi.is_safe = True
 
 register.filter('truncatehanzi', truncatehanzi)