Create a script to fix Taxonomized values stored in the wrong language

[teolemon]

Problem

• Some problematic apps have stored axonomized values in the wrong language, like fr:Dairy drinks or en:Boissons lactées
• https://world.openfoodfacts.org/category/en:boissons-lactees
• Create a script to fix Taxonomized values stored in the wrong language
• Since theoretically the loop is very long (140 languages), the algo could be directed on a specific tag like the one above, with a target value.

Part of

• 🧽 Data quality (tracker) #5538

[benbenben2]

I created a script for that.
#10711
Currently testing it in net environment.

Also see:
#7838

[github-throwaway]

@benbenben2 Whats the status on this? :) Anything where one can help?

[benbenben2]

I stopped working on it.
Main challenge is the new limitation on page requests per minutes.
We need to load many pages to retrieve all unknown tags and also to retrieve all products having this tag.
Here is an old script version written before this API limitation: https://github.com/openfoodfacts/openfoodfacts-server/pull/9581/files. Now the script is obsolete and I made a PR to remove it from the repo.

There are suggestions from the community to deal with it beside the API limitation on Slack: https://openfoodfacts.slack.com/archives/C02LDQDDD/p1725200136159869.
Python SDK is also facing this limitation: openfoodfacts/openfoodfacts-python#292
Downloading the database locally and use something like DuckDB might be an option but we will have to scroll on all products and process all tags to recreate an equivalent of the API for unknown tags. Which seems adding too much complexity.

I did not have further ideas so I put it aside for now.