Advertisement
Guest User

Untitled

a guest
Mar 19th, 2020
111
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
Python 0.72 KB | None | 0 0
  1. from docx.document import Document
  2. from docx.oxml.table import CT_Tbl
  3. from docx.oxml.text.paragraph import CT_P
  4. from docx.table import _Cell, Table
  5. from docx.text.paragraph import Paragraph
  6.  
  7.  
  8. def iter_block_items(parent):
  9.     if isinstance(parent, Document):
  10.         parent_elm = parent.element.body
  11.     elif isinstance(parent, _Cell):
  12.         parent_elm = parent._tc
  13.     else:
  14.         raise ValueError("something's not right")
  15.  
  16.     for child in parent_elm.iterchildren():
  17.         if isinstance(child, CT_P):
  18.             yield Paragraph(child, parent)
  19.         elif isinstance(child, CT_Tbl):
  20.             yield Table(child, parent)
  21.  
  22.  
  23. doc = Document('doc1.docx')
  24. result = [item for item in iter_block_items(doc)]
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement