pip install Spire.Doc

from turtle import st
from spire.doc import *
from spire.doc.common import *

def WriteAllText(fname:str,text:List[str]):
        fp = open(fname,"w")
        for s in text:
            fp.write(s)
        fp.close()

inputFile = "示例.docx"
outputFile =  "獲取的文本.txt"
     
#創(chuàng)建Document的對(duì)象
document = Document()

#載入Word文檔
document.LoadFromFile(inputFile)

#獲取文檔中的文本
text = document.GetText()

#將文本寫(xiě)入文本文件
WriteAllText(outputFile, text)
document.Close()

import queue
from spire.doc import *
from spire.doc.common import *
import os

outputPath = "Images/"
inputFile = "示例.docx"

if not os.path.exists(outputPath):
    os.makedirs(outputPath)

#創(chuàng)建Document的對(duì)象
document = Document()

#載入Word文檔
document.LoadFromFile(inputFile)

#創(chuàng)建一個(gè)隊(duì)列并將文檔元素放入其中
nodes = queue.Queue()
nodes.put(document)

#創(chuàng)建一個(gè)列表
images = []

#循環(huán)遍歷文檔元素
while nodes.qsize() > 0:
    node = nodes.get()
    for i in range(node.ChildObjects.Count):
        #獲取文檔元素的子對(duì)象
        child = node.ChildObjects.get_Item(i)
        #判斷子對(duì)象是否為圖片
        if child.DocumentObjectType == DocumentObjectType.Picture:
            picture = child if isinstance(child, DocPicture) else None
            dataBytes = picture.ImageBytes
            #添加到列表中
            images.append(dataBytes)
        #判斷子對(duì)象是否為復(fù)合對(duì)象
        elif isinstance(child, ICompositeObject):
            #添加到隊(duì)列中
            nodes.put(child if isinstance(child, ICompositeObject) else None)

#保存圖片
for i, item in enumerate(images):
    fileName = "Image-{}.png".format(i)
    with open(outputPath+fileName,'wb') as imageFile:
        imageFile.write(item)
document.Close()