Recipes imported through JSON does not have images and the Nutrional values are not imported.

[akshara-tg]

First Check

• I used the GitHub search to find a similar requests and didn't find it.
• Checked the tasks tagged issues and verified my feature is not covered

Please provide a concise description of the problem that would be addressed by this feature.

This feature will help users to address the limitations with the recipes imported through JSON method.

Please provide a concise description of the feature that would resolve your issue.

I use LLM models to generate recipes in JSON format:
e.g., prompt:

I prepare milk Kefir at home. I have a small blender that I use to prepare smoothies.
Can you suggest drink recipes that I can make with milk Kefir.
Usually I have these fruits, vegetables & other ingredients at home: Banana, Apple, Sapota, Carrot, Beetroot, Dates, Cashew, Almonds, Honey, Pink Salt, Oats.
Also provide each recipe in JSON format which is in compliant with Schema.org Recipe Type.
Include the Nutrional and caloric details in the recipe.

Issue:1

When I import this into the app, there will be no images,

Can you please provide an option to generate the image through OpenAI integration? The LLM model shall look at the recipe details and generate an appropriate image and that shall be attached to the recipe.

Issue:2

The JSON contains the Nutrional details, but on importing this recipe it is not available within the app.

{
  "@context": "https://schema.org",
  "@type": "Recipe",
  "name": "Banana Milk Kefir Smoothie",
  "recipeIngredient": [
    "1 cup milk kefir",
    "1 banana",
    "1 tablespoon honey",
    "1 tablespoon oats",
    "A pinch of pink salt"
  ],
  "recipeInstructions": [
    "Add the milk kefir, banana, honey, oats, and pink salt to the blender.",
    "Blend until smooth.",
    "Pour into a glass and enjoy!"
  ],
  "nutrition": {
    "@type": "NutritionInformation",
    "calories": "300",
    "proteinContent": "10g",
    "fatContent": "3g",
    "carbohydrateContent": "58g",
    "fiberContent": "4g",
    "sugarContent": "30g"
  }
}

Please consider and list out some caveats or tradeoffs made in your design decision

May be a better approach is to incorporate this entire thing as a new import feature. What I mean is; User shall be able to type in the LLM prompt and mealie app shall generate the recipe, images and import that into the app.

Additional Information

• If this is accepted I'm willing to submit a PR to provide this feature
• If this is accepted I'm willing to help maintain this feature
• I'm willing to sponsor/pay a developer to do this work

[Kuchenpirat]

Issue 1

I don't think adding AI-generated images to Mealie directly is necessary. There are problems with reliability for more complex recipes, and finding high-quality images of nearly everything is simple and fast. Relying on AI (which could potentially get it wrong) for recipe images doesn't seem justified.

If the JSON data contains a link to an image, as seen in the schema present on most recipe pages, this will also be scraped and added to the recipe.

Issue 2

I tried the JSON linked above, and it included all nutritional information.
This might be an issue of not having nutritional values enabled for the household or the recipe. These can be enabled via the household settings for default behaviour or the recipe settings, respectively.

[akshara-tg]

@Kuchenpirat
Thank you for the swift response.

I agree with you regarding Issue # 1.

I was not aware that nutritional values are a setting. It worked fine after I have enabled that and re-imported the recipe.
Is there any reason that is not enabled by default?

[Kuchenpirat]

All the data is scraped and saved so you don't have to Reimport recipes.
I am not sure, maybe not to have too many element on the recipe page.