pyhton Office文档转txt

最新推荐文章于 2025-09-22 15:48:21 发布

原创

最新推荐文章于 2025-09-22 15:48:21 发布 · 316 阅读

1 ·

CC 4.0 BY-SA版权

文章标签：

#python

只能说是网上代码进行优化后的究极缝合怪

import sys
#from win32com.client import Dispatch, constants
from pptx import Presentation
from pptx.util import Cm, Pt
import codecs
import pandas as pd
import win32com.client
import textract
import docx2txt
import docx
def change_doc_to_txt(word_path, save_path):
    word = win32com.client.Dispatch('Word.Application')  # 调用word应用
    doc = word.Documents.Open(word_path)
    print('保存中。。。')
    doc.SaveAs(save_path, 2)  # 保存格式为txt
    doc.Close()
    word.Quit()
def change_docx_to_txt(word_path, save_path):
    print('读取中。。。')
    doc = docx.Document(word_path)
    f = open(save_path, "w")
    for paragraph in doc.paragraphs:
        f.write(paragraph.text)

    for table in doc.tables:
        for row in table.rows:
            for cell in row.cells:
                f.write(cell.text)
    f.close()

def change_ppt_to_txt(word_path, save_path):
    #兼容ppt和pptx
    # ppt = win32com.client.Dispatch('PowerPoint.Application')
    # pptSel = ppt.Presen