合并两个文件内容的python脚本

MergeContent.py

# -*- coding: utf8 -*-
#!/usr/bin/python
import os
import re
import codecs
import sys
import tempfile
import string

reload(sys)
sys.setdefaultencoding('utf-8')  # 允许打印unicode字符

content_dir = 'C:\Users\Administrator\Desktop\docs'
meta_dir = "E:\\git\\cloudchou.github.io\\_posts"


def get_file_content(filename):
    f = codecs.open(filename, 'r', 'utf-8')
    s = f.readlines()
    f.close()
    return s


def write_file_content(filename, content, append):
    if append:
        f = codecs.open(filename, 'a+', 'utf-8')
    else:
        f = codecs.open(filename, 'w', 'utf-8')
    f.writelines(content)
    f.close()

files = os.listdir(meta_dir)

for filename in files:
    filename = unicode(filename, 'gbk')
    meta_file_path = unicode(meta_dir, 'gbk') + "\\" + filename
    meta_file_content = get_file_content(meta_file_path)
    meta_file_content = re.sub(
        r'(---.*---).*', r'\1', string.join(meta_file_content,''), flags=re.M | re.S)
    tmpfd, tempfilename = tempfile.mkstemp()
    write_file_content(tempfilename, meta_file_content+"\r\n", False)

    content_file_name = re.sub(
        r'(\d{4}-\d{2}-\d{2})-(.*)\.md', r'\2'+'.md', filename)
    print content_file_name
    content_file_path = unicode(content_dir, 'gbk') + "\\" + content_file_name
    if not os.path.exists(content_file_path):
        print content_file_name+" can not be done"
        continue
    content_file_content = get_file_content(content_file_path)
    content_file_content = string.replace(
        string.join(content_file_content,''), "\n", "\r\n")
    write_file_content(tempfilename, content_file_content, True)

    # os.remove(meta_file_path)
    # os.rename(tempfilename, "D:\1.txt")
    tmp_file_content = get_file_content(tempfilename)
    write_file_content(meta_file_path, tmp_file_content, False)

    # newFileName, _ = re.subn(
    #     r'(\d{4}-\d{2}-\d{2})-.*\.md', r'\1-'+'.md', filename)
    # print newFileName
    # break