Config proposal (API Simplification)

Start by reviewing issue #83 and the current SDK configuration/API entry points; this proposal introduces Feature, Category, Text, Number, and Schema. Compare the existing API with the shown JSON output and clarify the intended simplification with maintainers. Done means the design is agreed and its implementation scope is defined.

This is number 3 from #83 , @mhaas , what do you think?

import json
class Feature:
    def __init__(self,columns:list=[]):
        self.columns = columns
        self.repr = self._get_def()
    @property
    def featuretype(self):
        pass
    def _get_def(self):
        cols = []
        for col in self.columns:
            cols.append({"label": col, "type": self.featuretype})
        return cols
    def __add__(self,other):
        self.repr=self.repr+other.repr
        return self
    def __repr__(self):
        return str(self.repr)

class Category(Feature):
    def __init__(self,columns:list=[]):
        super(Category,self).__init__(columns)
    @property
    def featuretype(self):
        return "CATEGORY"
    
class Text(Feature):
    def __init__(self,columns:list=[]):
        super(Text,self).__init__(columns)
    @property
    def featuretype(self):
        return "TEXT"

class Number(Feature):
    def __init__(self,columns:list=[]):
        super(Number,self).__init__(columns)
    @property
    def featuretype(self):
        return "NUMBER"

class Schema:
    def __init__(self,features=None,labels=None,name=None):
        self.features = features
        self.labels = labels
        self.name = name
        self.repr = self._get_def()
    def _get_def(self):
        return json.dumps({
                            "features": self.features.repr,
                            "labels": self.labels.repr,
                            "name": self.name,
                            },indent=2)
    def __repr__(self):
        return str(self.repr)

features = Text(['MYTEXT'])+Number(['ANumeric'])+Category(['Manufacturee','description'])
labels = Category(['Label1','Label2'])+Number(['Label3'])
schema = Schema(features,labels,"bestbuy-category-prediction")
print(schema)

result:
{
"features": [
{
"label": "MYTEXT",
"type": "TEXT"
},
{
"label": "ANumeric",
"type": "NUMBER"
},
{
"label": "Manufacturee",
"type": "CATEGORY"
},
{
"label": "description",
"type": "CATEGORY"
}
],
"labels": [
{
"label": "Label1",
"type": "CATEGORY"
},
{
"label": "Label2",
"type": "CATEGORY"
}
],
"name": "bestbuy-category-prediction"
}