How to split a document after each heading?

[baranskib]

Hello,

I have multiple docx files in a folder and each of the docx has a couple of articles in it. The articles each have a heading and normal text. What I am trying to do is to iterate through all of those files and split each docx into separate text files. So if my original file1.docx has 4 articles, I want it to be split into 4 separate files each with its heading and text.

I got the part where it iterates through all of the files in a path and I can read the headings and text separately , but I can't seem to figure out a way how to merge it all and split it into separate files each with the heading and the text.

Any help will be appreciated, I hope that someone has already done something similar.

Thanks

[baranskib]

This is what it looks like, so I want to grab and separate the paragraphs between the headings.

[skurzinz]

@baranskib I see you closed this a while ago, but would you mind posting the solution as well? There might be people who would consider this a useful post ;-)

[baranskib]

The solution was posted here: https://stackoverflow.com/questions/54409495/splitting-a-docx-by-headings-into-separate-files-in-python

[skurzinz]

Thanks @baranskib -- without wanting to avoid actual work, can I ask again for

You'll need an implementation of is_heading() and create_document_from_paragraphs().

The first one is doable, but for the latter, I am struggling with

def create_document_from_paragraphs(paragraphs, nbr):
    outdoc = docx.Document()
    for p in paragraphs:
        # outdoc.add_paragraph(p) # returns type error
        outdoc.add_paragraph() # obviously only adds an empty paragraph
    outdoc.save(str(nbr) + '.docx')

nbr = '02' # nbr is a parameter for labelling the resulting split files

create_document_from_paragraphs(paragraphs, nbr)

This is becoming off-topic for this issue and more python support, but as it's related to the original topic, I still dare ask here...

[baranskib]

import glob
from docx import Document


def is_heading(paragraph):
    """Checks if a paragraph is a heading.

    :param paragraph:
    :return:
    """
    if paragraph.style.name.startswith('Heading'):
        return True


def iterate_document_sections(document):
    """For each headed section, generate a sequence of paragraphs.
    Each sequence starts with a headed paragraph, followed by text paragraphs.

    :param document: Current Comet Issue
    :return: an article
    """
    paragraphs = [document.paragraphs[0]]
    for paragraph in document.paragraphs[1:]:
        if is_heading(paragraph):
            yield paragraphs
            paragraphs = [paragraph]
            continue
        paragraphs.append(paragraph)
    yield paragraphs


def create_document_from_paragraphs(paragraphs):
    """Iterates through the paragraphs containing articles and splits them into separate files.

    :param paragraphs: Article text
    :return: New document with a single article.
    """
    new_doc = Document()
    for counter, words in enumerate(paragraphs):
        new_content = words.text
        new_doc.add_paragraph(new_content)
        print (new_content)
    new_doc.save('Articles/new_doc' + str(counter) + '.docx')


for name in glob.glob('Word Comet/*.docx'):
    document = Document(name)
    for paragraphs in iterate_document_sections(document):
        create_document_from_paragraphs(paragraphs)

I found this. I think that's all the code that I used for this issue.

[baranskib]

Sorry about the formatting, but I guess it's readable enough.

[skurzinz]

Thanks a lot -- the code is legible enough (and thanks to @scanny for the cleanup)!
I'll try this way… In general, I wanted to avoid having to cough up the text in single paragraphs with their own formatting, as in my case the paragraph styles between the headings are still important; so I'll have to save the styles as well…

[scanny]

GitHub will format your code nicely if you surround it with triple-backticks (separate lines above and below). Follow the top bacticks with "python" and you get the language-specific highlighting. I adjusted your post above, you can just open that up in edit mode to see an example.