doc文件转txt

doc文件转txt

# -*- coding:utf-8 -*-
# 安装pywin32包 http://sourceforge.net/projects/pywin32/files/pywin32/
# windows 7下使用通过
#

import os, sys
from fnmatch import fnmatch
import win32com.client

if len(sys.argv)<=2:
    print "python %s inputdir  outputdir" % os.path.basename(sys.argv[1])
    sys.exit(1)

input = sys.argv[1]
output = sys.argv[2]

if not os.path.exists(output): os.mkdir(output)

i = 1
wordapp = win32com.client.gencache.EnsureDispatch("Word.Application")
wordapp.Visible = 0
wordapp.DisplayAlerts = 0

for path, dirs, files in os.walk(input):
    docs = [os.path.abspath(os.path.join(path, f)) for f in files if fnmatch(f, '*.doc')]
    for doc in docs:
        doc2txt = os.path.join(output, os.path.basename(doc.rstrip('doc') + 'txt'))
        doc2txt = os.path.abspath(doc2txt)
        if os.path.exists(doc2txt): continue
        print " processing [%d] %s " % (i, doc)
        i +=1
        wddoc = wordapp.Documents.Open(doc)
        wddoc.SaveAs(doc2txt, FileFormat=win32com.client.constants.wdFormatTextLineBreaks)
        wddoc.Close()
wordapp.Quit()