There have been multiple accounts created with the sole purpose of posting advertisement posts or replies containing unsolicited advertising.

Accounts which solely post advertisements, or persistently post them may be terminated.

Any good options to prune out similar-but-variable bits of text from a large document?

Nursing student here!

So we get a shit load of reading assignments, and since everything’s digital nowadays, I’ve been leaning a lot on text-to-speech software that effectively converts reading assignments to listening assignments.

The problem is textbooks have a LOT of just… noise. Every image has something like “FIGURE 13.5 SURGICAL DISASTERS!” “FIGURE 13.6 YOU GOT SUMMONED TO COURT!” etc. In-text citations are EVERYWHERE, copyright info is EVERYWHERE… reading the content, you just skip over all that crap, but pasting it into a TTS service, all that trash gets spoken aloud and adds up to a huge time sink every chapter, and distracts from the actual lesson.

Googling it, the best I’ve been able to come up with is doing a find and replace in MS word for things like FIGURE **.*^13 with wildcards on and the replace field blank… but it’s not very consistent - sometimes it works, sometimes not. Same with nuking parenthesis and the text within with (*)

All that said, I’m wondering if I’m approaching this wrong by using MS word in the first place. Would be absolutely amazing if I could save all the commands on standby, then run them at the same time. By end of the school program, we’re talking like 100 chapters from multiple books, so anything that lets me just nuke huge batches of BS as quickly as possible and dive right into the listening would be a godsend.

Thanks all!!

LANA_DEL_KARENINA ,

The good news: there is a tool built to solve this exact problem: regular expressions (aka regex)

The bad news: regular expressions are famously frustrating to read and write

Depending on how badly you want the problem solved and how patient you are, using online resources to craft some regular expressions would be the ticket

Sterile_Technique OP ,
@Sterile_Technique@lemmy.world avatar

hmmm “famously frustrating”, presumably to people who know what they’re doing, very likely translates to “WAY outside of my skill level”. Worth some digging though, especially now that I have a keyword! Thank you!!

BearOfaTime ,

There are regex tutorials online, and you can test your regex there.

I’d say, since you’re learning, this could be an opportunity that may be useful later.

Just start with one relatively simple thing, like maybe copyright stuff. Work on getting regex to match that properly throughout a doc, and enjoy the improvement. Then when ready, tackle the next thing.

otter ,

IMHO, this is one of the applications wherein “AI” in its current form can really shine. Even the low monthly cost of GPT could be worth it if only to be able to train your own bot on specifics for your own use-case. Hell, there might even be one already made that’s close enough? If you’d like me to give a quick look, LMK. 🖖🏽

mozz ,
@mozz@mbin.grits.dev avatar

Copy paste into an LLM, piece by piece; ask it to keep the text the same but strip out the crap you don’t need. Give it specific examples of what things you want it to strip out. Then feed that output to the text-to-speech.

My recommendations in order of which to try first would be:

  • ChatGPT free tier
  • Claude.ai $20/mo tier
  • ChatGPT $20/mo tier

Re-give the instructions with every page; it won’t remember what to do. Honestly you probably want to start a new chat for each page / 2 pages / 5 pages or whatever. With simple instructions like this it can handle quite a lot of text at once so I wouldn’t be shy about how much you paste unless you start to hit limits or something.

Good luck, hope it helps

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • random
  • [email protected]
  • lifeLocal
  • goranko
  • All magazines
    •