解析word文档表格

发表于 
# -*- coding: utf-8 -*-
# from __future__ import print_function
import sys
reload(sys)
sys.setdefaultencoding('utf-8')

import win32com
from win32com.client import Dispatch, constants
w = win32com.client.Dispatch('Word.Application')
w.Visible = 1
w.Documents.Open( FileName = 'D:\muzi\\37.docx')  #路径不能存在中文
doc = w.ActiveDocument
count = doc.Tables.Count  #统计整个word文档有多少个表
a = doc.Tables(1).rows.Count # 返回表格的行数
b = doc.Tables(1).columns.Count # 返回表格的列数

import collections
Content = collections.OrderedDict()
for num in range (count):
    for i in range (doc.Tables(num+1).rows.Count):
        key = re.sub(r'\r\x07','',doc.Tables(num+1).Cell(Row= i+1,Column = 1).Range.Text) //正则化去除特殊符号
        value = re.sub(r'\r\07','',doc.Tables(num+1).Cell(Row=i+1,Column = 2).Range.Text) 
        Content[key] = value
    filename = 'D:\muzi\m'  
    import json
    with open(filename+'.json','a') as outfile:    //打开已存在的文件
        json.dump(Content,outfile,ensure_ascii = False) //字典内容写入json文件中
        outfile.write('\n')