python正则表达式提取中文字符串

要做国际化的版本，需要把中文字符串都提取出来翻译，写了这个python脚本。

#!/usr/bin/python
# -*- coding: UTF-8 -*

'''
##
# @file match-chinese.py
# @brief 利用正则表达式提取中文字符串
# @author Jesse
# @version 1.0
# @date 2009-11-20
'''

import os,string
import re

directory = "."
output = "chinese.txt"

def match_chinese(s, f, i):
    global fd_output
    r = re.compile('\"[^\"]*[\x80-\xff]{3}[^\"]*\"')
    s_match = r.findall(s)
    for c in s_match:
        str = "%s ( %d ): %s\n" % (f, i, c)
        fd_output.write(str)

def istextfile(filename, blocksize = 512):
    return istext(open(filename).read(blocksize))

def istext(s):
    if "\0" in s:
        return 0
    
    if not s:
        return 1

    text_characters = "".join(map(chr, range(32, 127)) + list("\n\r\t\b"))
    _null_trans = string.maketrans("", "")
    t = s.translate(_null_trans, text_characters)

    if len(t)/len(s) > 0.30:
        return 0
    return 1
    
def read_file(f):
    if not istextfile(f):
        print "%s is NOT a text file" % (f)
        return
    #if not re.match(r".*\.[c|h]$", f):
    # return

    i = 0
    fd = open(f,'r')
    buff = fd.readlines()
    for line in buff:
        i += 1
        match_chinese(line, f, i)
    fd.close()

def walk_dir(dir):
    try:
        if not os.path.exists(dir):
            print dir, ": No such file or directory."
            return
        if not os.path.isdir(dir):
            read_file(dir)
            return
        files = os.listdir(dir)
        for file in files:
            file_abs = dir + "/" + file
            if os.path.isdir(file_abs):
                print file_abs,"is ...DIR..."
                walk_dir(file_abs)
            else:
                read_file(file_abs)
    except:
        print "Exception occured"

def main():
    global fd_output
    fd_output = open(output, 'w')
    walk_dir(directory)
    fd_output.close()

if __name__ == "__main__":
    main()


参考资料：

编码相关：

http://blog.zol.com.cn/1356/article_1355079.html

ftp://ftp.astron.com/pub/file/file-5.03.tar.gz

python相关：

http://www.blogjava.net/Skynet/archive/2009/05/02/268628.html

http://iregex.org/blog/regex-to-match-chinese.html

http://hi.baidu.com/sinomazing/blog/item/cba18400f4473519738b6508.html

http://blog.donews.com/limodou/archive/2004/08/30/83538.aspx