Express Typed Query

Exhausted from double checking types passed to your data layer? Well this simple package leverages the power of qs and parses your query sting to usable types. Use real types when exchanging parameters without JSON serialization. Make sure your users can get the data they need in the mode flexible way possible. Converts query strings to Javascript typed objects including dates and malformed JSON

Install

npm

npm install express-typed-query

yarn

yarn add express-typed-query

pnpm

pnpm add express-typed-query

Usage

import express from 'express';
import eqs from 'express-typed-query';

const app = express();

eqs.configure(app, options);

import express from 'express';
import eqs from 'express-typed-query';

const app = express();

eqs(app, options);

import express from 'express';
import { configure } from 'express-typed-query';

const app = express();

configure(app, options);

const express = require('express');
const { configure } = require('express-typed-query');

const app = express();

configure(app, options);

const express = require('express');
const eqs = require('express-typed-query');

const app = express();

eqs.configure(app, options);

Options

Logger

This package uses JavaScript console logging by default. You can provide a log level to adjust what is logged using the log level option. The default level is "error" if no option is provided.

const options = { logging: { level: 'debug' } };

If you'd like to provide your own logger you can use the following. It must have the following levels or aliases in order of precedence:

error 0
warn  1
info  2
debug 3
trace 4

const options = { logging: { logger: <Logger> } };

Dates

If you'd like dates to be parsed from strings you can use the following option

const options = { dates: true };

This option will try parse date strings that are parsable into date objects.

Hail Mary

Sometimes complex JSON queries may have mistakes or incorrect quotes. Using the hailMary flag will try replace all quotes in a string that appears to contain JSON and attempt to parse it. This operation is risky as it makes assumptions regarding the string. If it fails an error will be thrown indicating it was unable to be parsed.

If the hailMary flag is not set the original string will be parsed back and error handling will be required by the caller.

const options = { hailMary: true };

Input
const qs = 'filter={string: "string", boolean: true, number: 1, float: 1.11, null: null, array: ["string", true, 1, 1.11, null] }';

Output
const output = { filter: {
  string: 'string',
  boolean: true,
  number: 1,
  float: 1.11,
  null: null,
  array: [
    'string',
    true,
    1,
    1.11
  ]
}}

Parameter Parsing

Utilizing terms from the Open API 3.1.0 spec, this module will parse form and deepObjects using the explode settings to convert each to their respective representation.

Repeated Keys keys=x&keys=y

Primitive types with repeated keys will be parsed to their associated type for instance

Input
'?id=1&id=2&name=John&name=Doe'

Output
{
  id: [1, 2],
  name: ["John", "Doe"]
}

Array Keys keys[]=x&key[]=y

This format ensures a key/value intended to be an array will be interpreted as such.

Input
'?id[]=1&name[]=John'

Output
{
  id: [1],
  name: ["John"]
}

If we use the repeated key syntax and the API is expecting an array as in the following example:

Input
'?id=1&name=John'

Output
{
  id: 1,
  name: "John"
}

It results in more processing by the route to handle the case where the data layer is expecting an array.

Deep Objects

Leveraging Open API terminology again, we can nest arrays of objects as follows

Input
'?user[a]={ "id": 1, name: "John" }'

Output
{ user: { a: { id: 1, name: 'John' } }}