0.1.4 • Published 5 years ago

povjs v0.1.4

Weekly downloads
3
License
MIT
Repository
github
Last release
5 years ago

pov

Point-of-view, everybody's got one!

Say you work in the presentation layer and the business concepts that make sense to you are foo and bar, this is your point-of-view.

However, you need to access/mutate these concepts through an API which exposes them as oof and rab.

The obvious solution is to use mappers to transform from one to the other for both read & write, but let's try another approach, that of having a live point-of-view.

const ipa = api.fetch();
// ipa === {oof: 1, rab: 2}

const my_concepts = my_point_of_view(ipa);
// my_concepts === {foo: 1, bar: 2}

Now we can both read & write oof as foo:

my_concepts.foo === ipa.oof; // true

my_concepts.foo = -1;

my_concepts.foo === ipa.oof; // still true

We need to provide some kind of mapping information to associate foo with oof, and bar with rab:

const my_mapping = { foo: 'oof', bar: 'rab' };
const my_concepts = my_point_of_view(ipa, my_mapping);

We don't just need to rename concepts but often times also apply some transformation(s) to fit our point-of-view.

For example API also returns a date of birth 1980-01-01T01:01:01.001Z that we want to use without the time part 1980-01-01.

So instead of only providing the name of the source concept (as oof for foo), our maping should define functions to get & set our concept:

const mapping = {
  foo: 'oof',
  bar: 'rab',
  birth: {
    get: _ipa => _ipa.birth.toISOString().slice(0, 10),
    set: (_ipa, _date) => {
      _ipa.birth = new Date(_date);
    },
  },
};

We can use these mapping functions to perform more complex transformations, including aggregation of concepts and accessing nested levels of concepts.

Since it is a common requirement, an array mapping with get & set functions is provided out-of-the-box. Here is a real example of use of pov:

import pov, { array } from 'povjs';

const ipa = {
  oof: 1,
};

// let's encapsulate the use of pov and its mapping
function mapFoo(source) {
  return pov(source, {
    foo: 'oof',
  });
}
const mapped = mapFoo(ipa); // mapped === {foo: 1}

// now the use of array
const ipas = {
  soof: [{ oof: 2 }],
};
function mapFoos(source) {
  return pov(source, {
    foos: array()('soof', mapFoo),
  });
}
const mappeds = mapFoos(ipas);
mappeds.foos[0].foo === ipas.soof[0].oof; // true

mappeds.foos[0].foo = 3; // true
mappeds.foos[0].foo === ipas.soof[0].oof; // still true

mappeds.foos = [{ foo: 4 }];
mappeds.foos[0].foo === ipas.soof[0].oof; // still true

All good, but why array()(...)?

Because array is a currying function which may accept two parameters, respectively getter and setter, both functions to provide pluggable extensions to the read & write abilities.

As a matter of fact, pov is also a currying function which also accepts the same number of parameters but for convenience, is exposed pre-configured with defaults.

import pov, { pov as currying_pov } from 'povjs';

Here currying_pov() and pov will be equivalent.

With these extension points, we can provide custom getter & setter as interceptors, for example:

import { pov as currying_pov, getter, setter } from 'povjs';

// decorate default getter & setter with...
function _getter(source, path) {
  console.log('getter', { path });
  return getter(source, path);
}
function _setter(source, path, value) {
  console.log('setter', { path, value });
  setter(source, path, value);
}
const pov = currying_pov(_getter, _setter);

Another example would be to provide custom getter & setter to support nested properties:

const ipa = {
  oof: {
    rab: 1,
  },
};
const mapped = pov(ipa, {
  bar: 'oof.rab',
});

This nested path access is such a common requirement, but to avoid reinventing the wheel, a povjs sister project povjs-lodash will provide these functionalities, along with its dependencies to a minimal required subset of the excellent lodash library.

0.1.4

5 years ago

0.1.3

5 years ago

0.1.2

5 years ago

0.1.1

5 years ago

0.1.0

5 years ago

0.0.6

5 years ago

0.0.5

5 years ago

0.0.4

5 years ago

0.0.3

5 years ago

0.0.2

5 years ago

0.0.1

5 years ago