Improve scraping of spisbedre.dk #1489
Conversation
| servings = int(self.recipe_json.get("serving_size")) | ||
| yield_type = ( | ||
| "item" if self.recipe_json.get("serving_size_unit_id") != 1 else "serving" | ||
| ) | ||
| return f"{servings} {yield_type}{'s' if servings != 1 else ''}" |
There was a problem hiding this comment.
| servings = int(self.recipe_json.get("serving_size")) | |
| yield_type = ( | |
| "item" if self.recipe_json.get("serving_size_unit_id") != 1 else "serving" | |
| ) | |
| return f"{servings} {yield_type}{'s' if servings != 1 else ''}" | |
| return self.schema.yields() |
We can gather the yield info directly from the json without any custom logic!
There was a problem hiding this comment.
That's only partly true as spisbedre.dk only renders a numeric value for the recipeYield tag, which doesn't indicate if the yield is per person or a number of items.
From the get_yields function in _utils.py I gathered that yield should reflect either servings or items, I hope this is a correct assumption.
So, to get the exact indication, it's necessary to fetch the information from the JSON structure which isn't in the JSON-LD format hence the additional parsing.
The third test case (mentioned below) is a recipe for buns, and the yield is number of buns - I hope this makes sense.
| "Fyld marengsen i en sprøjtepose, og sprøjt den ud i pæne toppe oven på hver bolle. Brænd marengsen med en lille gasbrænder, så den får gyldne kanter. Servér fastelavnsbollerne straks med en god skefuld lemoncurd på toppen." | ||
| ], | ||
| "category": "Desserter, Kager, Fastelavnsboller", | ||
| "yields": "12 items", |
There was a problem hiding this comment.
| "yields": "12 items", | |
| "yields": "12 servings", |
There was a problem hiding this comment.
Please see the comment above, regarding the yield formatting.
|
lgtm, thanks for the explanations |
Improve scraping by including nutrients, and equipment from the recipes.
Formatting of yields and ingredient units are also aligned with the website rendering.
Closes #1488