pdf Import Request #1588
Replies: 18 comments 25 replies
-
|
Is it even possible to add individual PDFs? |
Beta Was this translation helpful? Give feedback.
-
|
in addition import .doc based recipes, i have a huge collection on .pdf or .doc rexipes that i previously kept in evernote. ive been trying to import them into mealie, but copy pasting one by one is pretty tedious, it would be nice if there was a quick way to read in with basic achema looking for title, cook time and ingridients |
Beta Was this translation helpful? Give feedback.
-
|
In a previous version, it could read .PDF really well and you could import the text into the recipe. I guess it was removed? |
Beta Was this translation helpful? Give feedback.
-
|
Add PDF's to Mealie... Trying to keep this request alive |
Beta Was this translation helpful? Give feedback.
-
|
+1 If I could also add, there is a docker app called hp-to-scanner. It allows you send scans from your printer to your computer/docker container. This would be an awesome addition as it would mean that people have all these paper backups could run them through their scanner, and be automatically sent to mealie to import into the system. https://github.com/manuc66/node-hp-scan-to I use it currently with another application called paperless |
Beta Was this translation helpful? Give feedback.
-
|
My wife is harassing me over the missing import from PDF functionality. |
Beta Was this translation helpful? Give feedback.
-
|
lol, best motivation ever
…On Sat, Jul 20, 2024 at 12:15 PM lauris-nl ***@***.***> wrote:
My wife is harassing me over the missing import from PDF functionality.
It's not that I ever would harass you.
—
Reply to this email directly, view it on GitHub
<#1588 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ABY7PVIZKDL4CCR54Z246ETZNKLJTAVCNFSM6AAAAABJMEUQ56VHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTAMJQGMZTMOA>
.
You are receiving this because you commented.Message ID:
***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
I've tried to ChatGPT to mealie some pdf recipes and it is an absolute disaster! Please bring this feature in Mealie! |
Beta Was this translation helpful? Give feedback.
-
|
Another vote for this feature. Blue Apron is currently unable to be scraped due to bot detection but I could download the recipe card to pdf and import it manually as an easy work around. |
Beta Was this translation helpful? Give feedback.
-
|
i would love to the this option. me and my friends all have tons of pdf receips. |
Beta Was this translation helpful? Give feedback.
-
|
Would be great to have this functionality. It was a hit or miss option on Tandoor, so having a good implementation of this would be nice. Can i suggest that not trying to read or process text on the PDF would be ideal, and just display a folder or the uploaded pdf directly might be the best idea. Given how messy text processing can be on PDF files, and the. You throw images in and nutrient scraping etc. |
Beta Was this translation helpful? Give feedback.
-
|
My issue with Tandoor as well |
Beta Was this translation helpful? Give feedback.
-
|
The newest nightly has the ability to import via image using OpenAI: #3974; it can also optionally translate the recipe for you (with varying success depending on which model you use) While PDFs are not currently directly supported, you can convert it to an image before importing (along with any other input type) |
Beta Was this translation helpful? Give feedback.
-
|
I thumbs' up it... Request not fulfilled |
Beta Was this translation helpful? Give feedback.
-
|
Maybe it's me, but it seems counter-productive to have to convert a text-based PDF file to a bitmap image in order to parse the text back out via OCR. |
Beta Was this translation helpful? Give feedback.
-
|
We’re pretty happy with the image integration choice. Overall parsing PDFs is more or less an impossible problem. There are a variety of formats, so have parable text others are effectively images. It’s a huge problem set that no one on the core team is interested in diving into and we don’t really have the capacity to take on the ask. At the end of the day, the way this would probably work is to just cover the pdf to an image. That said, I think we can leave this up to the community to write tool around. Our API is documented, so if you want to build a cli that takes a directory of PDFs, converts them to images and uses the open api importer. The reality is we can’t do everything, we’ve decided this is outside of the scope of what we want. That’s why we have an API, so you can extend the work we’ve done. Happy to promote a tool that can handle this problem. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks for the reply. I respect that you have your priorities for this useful application, which may not coincide with my specific needs. I also have to agree that PDF is a pretty messed-up file format, so I get the lack of desire to open that can of worms. Sounds like it's time to check-out the API. Thanks for the continued development on this great program.. |
Beta Was this translation helpful? Give feedback.
-
|
@hay-kot understand that parsig is an impossible view - but why not just allow for a direct view/embed of the PDF? Seems like it would be an easier route than trying to parse. |
Beta Was this translation helpful? Give feedback.
-
|
As an experiment I tried grabbing a screen print from a site, enabled AI and uploaded the PNG. Worked perfectly |
Beta Was this translation helpful? Give feedback.
-
|
But would you consider doing that for 300+ recipes? Serious question. |
Beta Was this translation helpful? Give feedback.
-
|
I'm trying to import an image but I'm constantly getting a "Something went wrong" error followed by a lot of python-looking errors in docker terminal with (error on line here) type of errors. Any suggestions? |
Beta Was this translation helpful? Give feedback.
-
|
Post the errors from the console, otherwise it isn't possible to assist. |
Beta Was this translation helpful? Give feedback.
-
|
Beta Was this translation helpful? Give feedback.
-
|
Dang I need openAI pro don't I? |
Beta Was this translation helpful? Give feedback.
-
|
Yeah the free tier of OpenAI won't really work. While I fully understand this isn't an option for everyone, you can bump up your account to tier 1 with only $5, which grants you access to all models and substantially higher rate limits. OpenAI-compatible APIs are also an option, such as ollama, but obviously you need pretty beefy hardware for that (or you need to rent processing power). |
Beta Was this translation helpful? Give feedback.
-
|
This image upload feature is a game changer, the wife approval factor is excellent. I was able to take our library of pdf recipes and using fitz convert all of them to a temporary directory as PNGs and then load them into Mealie. The only ways it could be better are:
Well done @michael-genson and team! |
Beta Was this translation helpful? Give feedback.
-
Yup! Fully agree, here, that's definitely the next major step with this feature. I also want to support more import types (e.g. directly uploading PDFs rather than requiring you to convert it to an image first, or arbitrary text documents) also with bulk options.
Great question, I'll add this to my todo list. No clue if it's feasible or not, but that would be awesome. |
Beta Was this translation helpful? Give feedback.
-
|
I am on the current newest, v1.12.0, but do not see the option to upload via image. I do see in the config settings that openAI is not configured, but I don't see where to set that up. What are the steps to start using the AI image uploader? |
Beta Was this translation helpful? Give feedback.
-
|
That's where I'm looking and I only have the two original options.
|
Beta Was this translation helpful? Give feedback.
-
|
Can you share your docker compose, just to be sure? Reacting any sensitive info |
Beta Was this translation helpful? Give feedback.
-
environment:
- ALLOW_SIGNUP=true
- PUID=1000
- PGID=1000
- TZ=US/Mountain
- MAX_WORKERS=1
- WEB_CONCURRENCY=1
- BASE_URL=https://mealie.mydomain.com
- OPENAI_API_KEY=sk-proj-loooooooooongstriiiiiing
- OPENAI_MODEL=gpt-4o
- OPENAI_ENABLE_IMAGE_SERVICES=false
- OPENAI_WORKERS=2
- OPENAI_SEND_DATABASE_DATA=false
- OPENAI_REQUEST_TIMEOUT=60 |
Beta Was this translation helpful? Give feedback.
-
It looks like you have image services disabled, so you won't have the option to import a recipe via an image |
Beta Was this translation helpful? Give feedback.
-
|
Amazing. Yeah that would do it. |
Beta Was this translation helpful? Give feedback.
-
|
I would like to use this feature, but I want to use Azure openAI resource since I have credits with work. I tried to use it, however, for it to work in azure, we need to specify the api version as a query params: However, when I look at your code, you take the full given url and append /chat/completions at the end. This ends up with a call to: Would there be a way to specify, in environment variables, the api-version used? so that you can append it at the end without us having to go play in py files (we're not all software developers ;) ) |
Beta Was this translation helpful? Give feedback.
-
|
Kind of silly that Microsoft has OpenAI compatible endpoints, except for requiring additional params. Can you open an issue for this specifically? We should be able to support custom params to accommodate this |
Beta Was this translation helpful? Give feedback.
-
|
Yeah right! But you know, Microsoft will do Microsoft stuffs... I'll open a ticket and link it here once done. |
Beta Was this translation helpful? Give feedback.
-
It's common to have a repository of pdf recipies, a bulk import of pdfs would be a great feature addition!
Beta Was this translation helpful? Give feedback.
All reactions