Extensible Typed Schemas

Write simpler, typed schemas and share them with both your frontend(s) and backend(s).

Because the frontend needs to know more coming straight from the app's backend. Especially if it's reactive.

Introducing ETS:

Typed schemas with contexts, annotations, endless descriptivity and cross-references.

Schemas you can also communicate to your frontend, not only the backend.

Assuming the schemas are always meant to describe an API

How it works

• On parse (ETSParser) -> it generates a JSON-like data object containing all declarations & types;

• On compile (ETSCompile) -> the previously parsed object first gets transpiled to typescript (includes types and interfaces for each schema) and also handled to TS-Compiler.

This repository uses the two above (in order) for every schema file inside the defined datamodel dir. See configuration options -> (This can be changed by altering the schemas.config.ts file)

The parser

Generates full-tree, dynamically-typed TS files, ready to be feed in the ETSCompiler class which can output to multiple supported formats such as RxJsonSchema, JsonSchema

How it's supposed to work:

On the very first run, it scans all .ets files in datamodel dir (schemas should always be top-level only) then:

• it creates a main declaration TS file -> generated.d.ts
• it scans the entire datamodel dir just for top-level schemas
  • it outputs a TS enum containing the filenames of all those files.

    enum Taxonomies = { User, Post, Comment } // can be named anything else

It also saves references for each schema of another in memory.

/datamodel
  User.ets
  Post.ets
  Comment.ets

The parser returns, for every input file, a .ts file that contains types, interfaces and the returned object of ETSCreator type. (This can be customized to also output .d.ts separately)