# pip install pywin32com
# ppt太大会读取失败
import win32com
from win32com.client import Dispatch, constants
import os
path=r'C:UsersmzjDesktopppt'
def ppt_text(p):
ppt = win32com.client.Dispatch('PowerPoint.Application')
ppt.Visible = 1
pptSel = ppt.Presentations.Open(p)
win32com.client.gencache.EnsureDispatch('PowerPoint.Application')
slide_count = pptSel.Slides.Count
for i in range(1, slide_count + 1):
shape_count = pptSel.Slides(i).Shapes.Count
# print(shape_count)
for j in range(1, shape_count + 1):
if pptSel.Slides(i).Shapes(j).HasTextFrame:
s = pptSel.Slides(i).Shapes(j).TextFrame.TextRange.Text
with open(get_filename_ext(p)[0]+os.sep+get_filename_ext(p)[1]+'.txt', "a+",encoding='utf-8')as f:
f.write(s + "
")
ppt.Quit()
def get_filename_ext(filename):
file_path, temp_file_name = os.path.split(filename)
shot_name, extension = os.path.splitext(temp_file_name)
return file_path, shot_name, extension
if __name__ == '__main__':
for i in os.listdir(path):
if 'ppt' in i:
ppt_text(os.path.join(path,i))