hedy is an ORM that focuses on the relation part and produces pojo's as result. It's backend is plugable, so you can use anything that has a CRUD-like interface, be it a normal database, a restful api or even a file.
const hedy = require("hedy");
const simpleAdapter = {
get: () => thing,
};
const store = hedy(simpleAdapter);
const query = store("things");
const result = await query.load();
result === thing;So there are few things to unpack here.
The Adapter connects the query with the backend. It receives the generated query when load() is called. It may return any data.
The store is the initialized ORM with the adapter set to it. It is a function that creates queries when called. The only argument is the table name.
The query is a immutable data structure that has all relevant configurations to create the request to the backend via the adapter. It contains e. G. the table name, pagination, filter and columns.
It has some methods to change this configuration. Calling them always returns a new query instance, but never mutates the existing one.
Calling load the load method sends the constructed query to the respective adapter function.
Relations are handled a little different from what you might know from existing ORM. Relations are handled as connections between queries. Since every query can have it's own backend, relations can also be defined between different backends. So you might connect you postgres data to you payment API transparently.
There are 5 possible relations you can use
Source query has a foreign key to the target query
Source queries primary key is a foreign key of one target query row
Source queries primary key is a foreign key of multiple target query rows
Source queries row is connected to target query row through a third query
Same as hasOne but it merges the related object to the this object in the result