Importing Recipes

Your cookbook can now import from several apps, using a single, simple flow:

Max upload size follows BODY_SIZE_LIMIT in your .env. Increase it if large archives fail.

Other supported formats

These match the code in src/lib/utils/import/registry.js and follow the same “pick type → upload → import” flow.

Source	File(s) you upload	What the ZIP/CSV must contain	Notes
Paprika	`.paprikarecipes` or `.zip`	A `.paprikarecipes` file, or a ZIP with `*.paprikarecipe` members	Categories ignored; photos imported if present
Chowdown	`.zip`	`_recipes/.md` and optional `images/`	Front-matter + body parsed; local images imported
Mealie	`.zip`	`recipes/<dir>/.json` and `recipes/<dir>/images/`	Ingredients/steps normalized; images imported from archive
MyRecipeBox	`.csv`	Columns like `title, description, ingredients, instructions, notes, …`	Maps columns directly to your schema; no images
Nextcloud	`.zip`	`<dir>/recipe.json` and `<dir>/full.jpg` (use `full.jpg`)	JSON is schema.org-ish; image imported
PlanToEat	`.csv`	Official PTE export CSV	Downloads `Photo Url` if present
Recipe Keeper	`.zip`	`recipes.html` and `images/<uid>_*.jpg`	HTML parsed; photos matched by filename
Tandoor	`.zip`	Top-level numbered zips each with `recipe.json` and `image` (with or without extension)	Uses embedded image; images without extension are handled

All importers use the same map-driven pipeline. Fields are normalized into your Recipe schema:

name, description, ingredients, directions, notes
prep_time, cook_time, total_time
servings
rating
source and source_url
image_url/photo_url (if present)
created (converted to a Date when available)
Flags: on_favorites, on_grocery_list, is_pinned, in_trash (defaults to false)
is_public (from the checkbox)

Photos are imported when available:

From archives (Paprika, Mealie, Nextcloud, Recipe Keeper, Tandoor): saved locally and attached to each recipe.
From CSVs (PlanToEat): we try to download the Photo Url.

Categories/tags are ignored for now across all importers.