Replies: 2 comments 7 replies
-
Open-source OCR libraries are generally terrible and layouts too variable to automate. HTML isn't going to work either as the importer relies on structured data formats. Manually typing (or copy /paste using your OS's built-in OCR) is going to be your best bet. Note that you can paste a list of ingredients and it will convert it into amount/unit/good/note format. |
Beta Was this translation helpful? Give feedback.
0 replies
-
I wrote a script to convert images or recipes into a JSON Schema which is compatible with Tandoor using Chat GPT APIs. |
Beta Was this translation helpful? Give feedback.
7 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I have been using the external storage source for uploading individual PDF files and adding them as standalone recipes. This works Okay, but, not Fantastic. It is not exactly as seamless as I would like it to be. Namely, the lack of OCR (which I know is a huge development task) to Title, ingredients, or Instructions. In addition, Untare (albeit not Recipes specific) cannot show the PDF files that are uploaded.
Is there any recommendations / guides that do not necessarily incorporate only Recipes' own functionality to get to a complete/proper ready to use import?
So far I have tried on a cookbook (that does not have headers or bookmarks):
Unfortunately this does not work out and provide issues with images (makes sense since the file path is local), instructions do not show up, ingredients do not show up.
Does anyone have a common method of exporting PDFs for import (without using the external files storage source)?
Beta Was this translation helpful? Give feedback.
All reactions