Working with Excel and Word, Several Questions Regarding Find and Replace

[Brandon_Pickert]

Looking to loop through Word doc files in a folder and update, or "find and replace" a specific section of the text in the documents. This text is typically in the same spot and same format (ie: "March 1, 2023") but it can change month to month. Another example is the footer with a unique initial (ie: "User - MM").

I have tried using the "os" library and docxtpl library to create variables in the Word document ({{Date}}) but realized having variables is not the route to go because this would be looping through "live" documents, aka, non-template documents. Finding and replacing the mentioned date string might be easy, but doing the same for a footer that has initials may be too unique for a find and replace method. Is there another method besides creating variables or finding and replacing out there? If finding and replacing is the best route, what tips could I utilize to future-proof this automation?

Current code based on variables that will not work:

doc = DocxTemplate(f"C:~filepath~\\{filename}")
        context = {'Footer': input_footer, 'Effective_Date': input_effective_date}
        doc.render(context)

Screenshot attached.
Windows 10, python version 3.11.1

[DeadParrot]

Glad to see you've already tried the docxtpl template.

I take it you're looking for a way to repeatedly replace these sections of text over time, and that these are "living" documents?

A naive solution, still using docxtpl, would be to have the value that you replaced the {{ template }} tag with recorded, and "re-templatize" it before running the process again?

It looks like dealing with the XML structure of the Word document contents may be a more direct way to deal with this, though.

Have you looked at the xml and python-docx libs?

https://docs.python.org/3/library/xml.et...ttree.html

https://python-docx.readthedocs.io/en/la.../text.html

These are just a couple that came up on a brief search: https://search.brave.com/search?q=word+d...on+library

I've done this kind of thing before, but your initial solution worked for me, since the documents were rarely updated after they were templated by me, or if they were updated, they didn't need to have any automation done on them afterward.

You might have luck with something involving XML tags, or the "styles" feature of Word, I guess?

---

If you could parse a .docx as XML, then add custom tags or attributes to the section of text you want to update automatically, and then generate the .docx back again with the new XML in it, that could work, I guess?

Perhaps this would be a stepping stone to doing that:

https://stackoverflow.com/questions/4018...words-that

---

Might be worth having a look at this, too: https://search.brave.com/search?q=python...+documents

https://stackoverflow.com/questions/3477...0#55733040

https://stackoverflow.com/questions/5637...ython-docx

[DPaul]

Hi,
Not hijacking your thread, but have you also found a clever way to work with legacy ".doc" and not ".docx" documents?
I asked this here some time ago, but it seems to be "difficult".
thx,
Paul

[Brandon_Pickert]

Hi,
Not hijacking your thread, but have you also found a clever way to work with legacy ".doc" and not ".docx" documents?
I asked this here some time ago, but it seems to be "difficult".
thx,
Paul

I have not. I'm sort of shooting from the hip on this project.

[Brandon_Pickert]

Glad to see you've already tried the docxtpl template.

Thank you DeadParrot for your detailed response. Your "styles" comment lead me down a great path and helped me solve the case!

What I ended up doing is looping through each paragraph and the runs of those paragraphs and comparing the style to a list of input styles I have in an excel then replacing it with the excel style input column of my excel. So, in my excel I have two columns, Style and Input, if the run matches a style, then replace run with input. Including headers and footers was tricky business but I eventually got there. The libraries I am using are docx, pandas, and os. This method works a lot better than find and replace because the "find" text could change as people update the document, but the style doesn't change.