• Introductory article
• Main documentation
• Examples

If you find this useful, please don't forget to star ⭐️ the repo, as this will help to promote the project.
Follow me on Twitter and GitHub to keep updated about this project and others.

schm is a library for creating immutable, composable, parseable and validatable (yeah, many *ables) schemas in JavaScript and Node.js. That's highly inspired by functional programming paradigm.

Play with schm on RunKit (click on Clone and edit this document at the bottom and skip login if you want).

const schema = require('schm')

const userSchema = schema({
  name: String,
  age: {
    type: Number,
    min: 18,
  },
})

userSchema.parse({
  name: 'Haz',
  age: '27',
})

await userSchema.validate({
  name: 'Jeane',
  age: 10,
})

Output:

// parsed
{
  name: 'Haz',
  age: 27,
}

// validate error
[
  {
    param: 'age',
    value: 10,
    validator: 'min',
    min: 18,
    message: 'age must be greater than or equal 18',
  },
]

The way you declare the schema object is very similar to mongoose Schemas. So, refer to their docs to learn more.

schm repository is a monorepo managed by lerna. Click on package name to see specific documentation.

Package	Version	Description
schm		The main package
schm-computed		Adds computed parameters to schemas
schm-express		Express middlewares to handle querystring and response body
schm-koa		Koa middlewares to handle querystring and response body
schm-methods		Adds methods to parsed schema objects
schm-mongo		Parses values to MongoDB queries
schm-translate		Translates values keys to schema keys

When submitting an issue, put the related package between brackets in the title:

[methods] Something wrong is not right # related to schm-methods
[translate] Something right is not wrong # related to schm-translate
Something wrong is wrong # general issue

PRs are welcome. You should have a look at lerna to understand how this repository works.

After cloning the repository, run yarn. That will install all the project dependencies, including the packages ones.

Before submitting a PR:

1. Make sure to lint the code: yarn lint or lerna run lint;
2. Make sure tests are passing: yarn test or lerna run test;

MIT © Diego Haz