[POC] Return parsed ingredients (name, unit, quantity)

[lizozom]

Currently ingredients are returned as unprocessed strings. I'm proposing a change to the existing ingredients API to returned an array of objects with the following structure (using quantulum3):

[{
    'name': 'white sugar', 
    'quantity': 2.0, 
    'unit': 'tablespoon'
}]

We could also make this a non breaking change by adding an optional input parameter or separating this into two APIs.

---

This PR only updates a few tests to demonstrate the change, and if the community agrees, I can update all other tests\parsers and the documentation.

Curious to hear what you think!

[lizozom]

@hhursev would you be interested in this type of improvement for project? If so, I could work on it.

[hhursev]

Hey!

Interested in what quantulum3 -like package can do out of the box on top of our .ingredients() method! Does it need to have numpy, scipy, sklearn installed?

I feel like if you are happy with the results you should continue on this idea! I'm thinking the proper approach for us is:

1. If user installs the package with pip install recipe-scrapers

<scraper>.ingredients()   # returns our current. the unprocessed strings.

2. If user installs with pip install recipe-scrapers[extras] the

<scraper>.ingredients()  # would return what you are suggesting.

so in a sense what you are proposing won't be in the core package but may overwrite the default .ingredients() method depending on with what instructions the package was installed with.

[dragonpop76]

Personally I'd love something like this!

[anguswg-ucsb]

@hhursev @lizozom For what its worth, the ingredient_slicer package provides this functionality using only base python and python's standard library. @hhursev If you wanted to implement some sort of quantity/unit extraction from the ingredients() method, WITHOUT bringing on any new dependencies (besides the ingredient_slicer itself) then this would work really well. The package is thoroughly tested and works really well for extracting units and quantities from ingredient strings.

An example from the README:

import ingredient_slicer

slicer = ingredient_slicer.IngredientSlicer("2 (15-ounces) cans chickpeas, rinsed and drained")

slicer.to_json()

{   
    'ingredient': '2 (15-ounces) cans chickpeas, rinsed and drained', 
    'standardized_ingredient': '2 cans chickpeas, rinsed and drained', 
    'food': 'chickpeas', 

    # primary quantity and units
    'quantity': '30', 
    'unit': 'ounces', 
    'standardized_unit': 'ounce', 

    # any other secondary quantity and units found in the string
    'secondary_quantity': '2', 
    'secondary_unit': 'cans', 
    'standardized_secondary_unit': 'can', 

    'gram_weight': '850.49', 
    'prep': ['drained', 'rinsed'], 
    'size_modifiers': [], 
    'dimensions': [], 
    'is_required': True, 
    'parenthesis_content': ['15 ounce']
}

Note: I am the author of this package and I'm shilling it because it works better than any other open source solution I could find (there are other good ingredient parsers out there, but most of them require large additional dependencies) and it filled a big hole in my work.