Crudity

A production ready CRUD backend server and node http middleware, working with files/mongodb/mariadb/etc.

Quick start

Take less than 10 seconds.

npx crudity

Open a browser on http://localhost:3000/api: It works!.

You can start configuring a crudity.json file. See Configuration

Install

Global mode

npm i crudity -g

On a local project

npm i crudity

Usage

Command line

npx crudity

Using crudity in express

Javascript

const express = require("express");
const { createServer } = require("http");
const { crudity } = require("crudity");

const app = express();
const server = createServer(app);
app.use(
  "/api/articles",
  crudity(server, "articles", {
    pageSize: 100,
    hateoas: "body",
  })
);
server.listen(3333, () => {
  console.log(`Server started on port ${server.address().port}`);
});

Typescript

import express from "express";
import { createServer } from "http";
import { AddressInfo } from "net";
import { crudity } from "crudity";

const app = express();
const server = createServer(app);
app.use(
  "/api/articles",
  crudity(server, "articles", {
    pageSize: 100,
    hateoas: "body",
  })
);
server.listen(3333, () => {
  console.log(
    `Server started on port ${(server.address() as AddressInfo).port}`
  );
});

Out of scope

This module only focus on CRUD operations. It does not perform API management:

• validation (sync or async)
• sanitizing (xss, trim, etc.)
• authentication / authorization
• logging (access log, trace)
• CORS checking

However, it offers some async validation operators:

• unique: avoid duplicate entry on some fields.
• that's it for the time being.

Note that you can write your own async validation operators using the CRUDService crudityRouter.service.

You can check these projects:

• Synchrone validation:
  • superstruct
• Authentication with OAuth2:
  • @jlguenego/express-oauth2-client
• logging access log
  • morgan

Crudity options

The middleware crudity(server: http.Server, resourceName: string, options: CrudityOptions) has following options called CrudityOptions:

• pageSize: number - default page size for retrieve all requests. 20 by default.
• storage - an object for specifying how to store the resource (ex: json file storage, mongodb storage). See [Storage options](# Storage options)
• hateoas: body, header, or none. Give hateoas informations through the HTTP request body, or the HTTP request header, or do not give hateoas information. Hateoas are kind of interesting related links (ex: next, previous, first, last page).
• delay: if you want to add slowness to the request (for debugging purpose)
• enableLogs: if you want to see logs.

Default options are:

const defaultOptions: CrudityOptions = {
  pageSize: 20,
  hateoas: "header",
  storage: {
    type: "file",
    dataDir: "./data",
  },
  delay: 0,
  enableLogs: false,
  validators: [],
};

Storage options

The storage object always has a type that can be:

• file
• mongo
• mariadb
• other value if implemented.

File (Json)

• dataDir: the directory where to store the json file.

Mongo

• uri: The mongo URI to connect to. See the MongoDB doc how to add options to it.

Other

You can implement your own storage service.

Playing with Crudity

With an HTTP Client you can call the server following this:

Create

Create one

POST /ws/articles HTTP/1.1
Content-Type: application/json

{"name":"Screwdriver","price": 2.99,"qty":100}

Create bulk

POST /ws/articles HTTP/1.1
Content-Type: application/json

[{"name":"Screwdriver","price": 2.99,"qty":100},{"name":"Hammer","price": 1.50,"qty":80}]

Retrieve

Retrieve all. Pagination applies.

GET /ws/articles HTTP/1.1

Retrieve the third page with default page size.

GET /ws/articles?page=3 HTTP/1.1

Retrieve the second page, with 50 items per page, reverse order by name, and then ordered by price ascending, finally filter by the first address city which must start by 'A' or 'a' (case insensitive).

GET /ws/articles?page=2&pageSize=50&orderBy=-name,+price&filter[addresses][0][city]=/a.*/i HTTP/1.1

The query string uses the qs node module format, integrated withing express. Crudity query options are:

• page: page index starting at 1. Default is 1.
• pageSize: number of items per page. Default is 20. 0 means no limit.
• orderBy: order by field name.
  • It is a concatenated string of fields that are to be ordered, separated by a comma character ,.
  • Each field can be reversed by prefixing it with -. + is also a optional prefix for ascending.
• filter: filter on field name given a javascript regex to be matched or simply a string to be equal to. The filter key must be an object with the same interface as the resource. The value are the regex or string wanted on the fields.
• select: list of fields to be returned, comma separated. * to return all. Default is *.

Retrieve one

GET /ws/articles/1234 HTTP/1.1

Update

PUT /ws/articles/1234 HTTP/1.1
Content-Type: application/json

{"id":"1234","name":"Screwdriver","price": 2.99,"qty":100}

PATCH /ws/articles/1234 HTTP/1.1
Content-Type: application/json

{"name":"Screwdriver"}

PATCH /ws/articles HTTP/1.1
Content-Type: application/json

{"qty":"100"}

Delete

Delete Many

DELETE /ws/articles HTTP/1.1
Content-Type: application/json

["1234"]

Delete All

DELETE /ws/articles HTTP/1.1

HATEOAS

Hateoas information is provided or not, according the 3 following modes:

• none: no hateoas is provided.

• header: the Link header is used (default).

• body: the returned body is an object containing two properties:

  • links: contains an array of all related hateoas links.
  • result: contains the normal result of the request.

You can configure in the crudity options the hateoas, but also overwrite the options directly in the request, using the query parameter hateoas or the HTTP header. The query string parameter has priority over the header.

Hateoas in query parameters

• ?hateoas=none: No hateoas info.
• ?hateoas=header: Hateoas info in the HTTP response header.
• ?hateoas=body: Hateoas info in the HTTP response body.

Hateoas in HTTP headers

• X-Crudity-Hateoas: none: No Hateoas info produced (default).
• X-Crudity-Hateoas: header: Hateoas info produced under the HTTP response header X-Crudity-Link in a JSON format: .
• X-Crudity-Hateoas: body: Hateoas info produced under the body response in a JSON format. Warning, the result of the request will be under the result key.

Example of request with HATEOAS in the body:

GET /ws/articles/1234 HTTP/1.1
X-Crudity-Hateoas: body

Response:

HTTP/1.1 200 OK
Content-Type: application/json
...

{"result": {"id":1234, "name": "Pliers", "price": 1.50, "qty": 300},
"links": ["next": "/ws/articles/1235", "previous": "/ws/articles/1233"]}

Configuration

Crudity can read a crudity.json file (or a javascript crudity.js file).

This $schema file can be used to get automatic completion in some IDE (VSCode, etc.) for the JSON file.

Example of crudity.json file:

{
  "$schema": "https://raw.githubusercontent.com/jlguenego/crudity/master/schema/crudity.json",
  "port": 3500,
  "publicDir": "./public",
  "cors": true,
  "resources": {
    "articles": {
      "pageSize": 10,
      "delay": 500
    },
    "users": {
      "pageSize": 15,
      "storage": {
        "type": "mongodb",
        "uri": "mongodb://localhost/crudity"
      }
    }
  },
  "rootEndPoint": "/api"
}

Open API

For a given resource, you can add the suffix /openapi.yml or /openapi.json to get the Open API specification document regarding the Crudity API.

Example:

http://localhost:3500/api/articles/openapi.yml

TODO

• Doc for implementing a new service
• Doc for integrating in express
• Exemple with validation, authentication, sanitizing, etc.

Thanks

Thanks to some url that helped me.

• https://www.moesif.com/blog/technical/api-design/REST-API-Design-Filtering-Sorting-and-Pagination/

Participating

Do not hesitate to bring your contribution to this project. Fork and Pull Request are welcome.

License

ISC

Author

Jean-Louis GUENEGO [email protected]