Valify was created to easily validate data structures. With a simple syntax it is ideal in many contexts for example in REST API

Documentation

• Installation
  • Browser
• Basic usage
• Model options
• Field options
• Error object
• Default values
• Nested models
• Promises
• Detect unknown fields
• Auto cast
• Manipulate data
• Immutability
• Undefined values
• Define custom types
  • Use multiple rules together
  • Arguments
• Locale
• Types
• Upgrade to V4

Installation

npm install --save valify

Browser

<script src="https://unpkg.com/valify/dist/valify.min.js"></script>

Basic usage

const Valify = require('valify');

// Define a model
const userModel = new Valify({
    firstName: 'string',
    lastName: 'string',
    age: 'int?', // this is not required
    role: {
        type: 'string',
        default: 'editor'
    },
    colors: ['string'],
    createdAt: {
        type: 'date',
        default: new Date()
    }
});

// A data object
const data = {
    firstName: 'Mike',
    lastName: 'Ricali',
    role: 'owner',
    colors: ['red', 'yellow', 'orange']
};

// Validate userModel
try {
    userModel(data);
} catch(e) {
    console.log(e.message, e.fields);
}

Model options

|Property|Type|Default|Description| |-|-|-|-| |usePromise|boolean|false|If you need to use with Promise must just add usePromise to model settings. Details| |detectUnknown|boolean|false|If you need to define a strict model where all the fields correspond to those defined, you can set detectUnknown to true. Details| |autoCast|boolean|false|Sometimes you may need to cast a string (where possible) to a primitive type. You can set autoCast to true. Details| |returnImmutable|boolean|false|Valify model returns also the data that you have passed for the validation, if you want an immutable data, set returnImmutable to true. Details| |overwriteUndefined|boolean|false|If you need manage undefined value with a default value, set overwriteUndefined to true, obviously works only if default is set. Details| |appendToError|object|{}|If you need to add custom properties to error stack|

Field options

|Property|Type|Default|Description| |-|-|-|-| |type|object,array,string,function|null|Type of control| |required|boolean|true|Indicates if the field is required| |default|any|null|Default value| |allowNull|boolean|false|Allow null value, overwrites all checks| |allowEmpty|boolean|true|Allow empty value, works for string, array and object| |locale|object|object|An object that contains locale strings that overwrites those globals| |convert|function|null|A function to manipulate/conversion data| |onError|function|null|A function triggered when an check fails|

Error object

Valify in case of errors returns an object with 2 properties:

• message is the first error occurred
• fields is an array of all errors occurred

{
    message: '"aParam.other.lastName" is required',
    fields: [
        {
            path: 'aParam.other.lastName', 
            message: '"aParam.other.lastName" is required', 
            field: 'lastName',
            type: 'string'
        }
    ]
}

Default values

You can set a default value for each field, this setting overwrites required property to false.

const Valify = require('valify');

// Define a model
const userModel = new Valify({
    name: 'string',
    role: {
        type: 'string',
        default: 'editor'
    }
});

// A data object
const data = {
    name: 'Mike Ricali'
};

try {
    userModel(data); //=> {name: 'Mike Ricali'}
} catch(e) {
    console.log(e.message, e.fields);
}

Nested models

It's possible also add nested model, for example you could have an array field like below:

const userModel = new Valify({
    firstName: 'string',
    lastName: 'string',
    records: [
        new Valify({
            id: 'int',
            accessOn: 'date',
            otherNested: new Valify({
                color: 'string'
            })
        })
    ]
});

// A data object
const data = {
    firstName: 'Mike',
    lastName: 'Ricali',
    records: [
        {
            id: 1,
            accessOn: '2017-12-23T00:01:00',
            otherNested: {
                color: 'red'
            }
        },
        {
            id: 2,
            accessOn: '2017-12-23T00:02:00',
            otherNested: {
                color: 'yellow'
            }
        },
        {
            id: 3,
            accessOn: '2017-12-23T00:03:00',
            otherNested: {
                color: 'green'
            }
        }
    ]
};

// Validate userModel
try {
    userModel(data);
} catch(e) {
    console.log(e.message, e.fields);
}

Using promise

If you need to use with Promise must just add usePromise to model settings.

// Define a model
const userModel = new Valify({
    firstName: {
        type: 'string',
        required: true
    },
    lastName: {
        type: 'string',
        required: true
    }
}, {
    usePromise: true
});

// A data object
const data = {
    firstName: 'Mike'
};

// Validate userModel
userModel(data).then(()=>{
    console.log('ok');
}).catch(e => {
    console.log(e);
    // An object like below
    /*
        {
            message: 'lastName is required',
            fields: [{field: 'lastName', message: 'lastName is required', path: 'lastName'}]
        }
     */
});

Detect unknown fields

If you need to define a strict model where all the fields correspond to those defined, you can set detectUnknown to true.

const userModel = new Valify({
    firstName: 'string',
    lastName: 'string',
    email: 'email'
}, {
    detectUnknown: true
});

try {
    userModel({
        firstName: 'Mike',
        lastName: 'Storm',
        email: '[email protected]',
        role: 'admin',
        age: 26,
    })
} catch (e) {
    console.log(e.message); //Unknown fields were detected: role, age
}

Auto cast

Sometimes you may need to cast a string (where possible) to a primitive type. You can set autoCast to true.

const userModel = new Valify({
    firstName: 'string',
    lastName: 'string',
    email: 'email',
    aBoolean: 'boolean',
    aNumber: 'number',
    aUndefined: 'undefined',
    aNull: 'null'
}, {
    autoCast: true
});

try {
    userModel({
        firstName: 'Mike',
        lastName: 'Storm',
        email: '[email protected]',
        role: 'admin',
        aBoolean: 'true',
        aNumber: '52',
        aUndefined: 'undefined',
        aNull: 'null'
    })
    //... done
} catch (e) {
    
}

Manipulate data

You may need to manipulate data before the validation.

// Define a model
const userModel = new Valify({
    firstName: {
        type: 'string',
        convert: value => value.toUpperCase()
    },
    lastName: {
        type: 'string',
        convert: value => value.toUpperCase()
    },
    age: {
        type: 'number',
        convert: value => parseInt(value)
    }
});

// A data object
const data = {
    firstName: 'Mike',
    lastName: 'Ricali',
    age: '25'
};

userModel(data);

console.log(data.firstName, data.lastName, data.age, typeof data.age); //=> MIKE RICALI 25 number

• Convert function returns:
  • value, current value
  • data, a copy of origin data object
  • be, a library used for several validations. More info on beJS

Immutability

Valify model returns also the data that you have passed for the validation, if you want an immutable data, set returnImmutable to true.

const userModel = new Valify({
    firstName: 'string',
    lastName: {
        type: 'string',
        convert: value => value.toUpperCase()
    },
    email: 'email'
}, {returnImmutable: true});

const data = {
    firstName: 'Mike',
    lastName: 'Storm',
    email: '[email protected]'
};

const newData = userModel(data);

console.log(data.lastName, newData.lastName);
//=> Storm, STORM

Undefined values

If you need manage undefined value with a default value, set overwriteUndefined to true, obviously works only if default is set.

const userModel = new Valify({
    aNumber: 'int',
    lastName: {
        type: 'string',
        default: 'Mike'
    }
},{
    overwriteUndefined: true
});

const a = ['hello'];

try {
    userModel({
        aNumber: 24,
        lastName: a[1] //=> index at 1 is undefined but will be applied default value "mike"
    });
    done();
} catch (e) {}

Define custom type

There are different ways to define custom types:

1) Globally, using static method addType or addTypes if you want add more than one type

Valify.addType('mycustom1', (value, data) => {
    console.log(data);
    return value === 10;
});

// it's also possible returns a string as error like below
Valify.addType('mycustom2', (value) => {
    if (value !== 10)
        return 'ops... must be 10'
});

// One method to add several types
Valify.addTypes([
    {
        name: 'mycustom3',
        fn: value => value === 'hello'
    },
    {
        name: 'mycustom4',
        fn: value => value === 'world'
    }
]);

// Define a model
const userModel = new Valify({
    aNumber: 'mycustom1',
    otherNumber: 'mycustom2'
});

// A data object
const data = {
    aNumber: 9,
    otherNumber: 11,
};

try {
    userModel(data);
} catch(e) {
    console.log(e.message, e.fields);
} 

2) Local, passing a function to type param

// Define a model
const userModel = new Valify({
    aString: {
        type: value => typeof value === 'string'
    },
    // or 
    aBoolean: value => typeof value === 'boolean'
});

// A data object
const data = {
    aString: 'hello',
    aBoolean: 5
};

try {
    userModel(data);
} catch(e) {
    console.log(e.message, e.fields);
} 

Use multiple rules together

If you need to define multiple checks in one type, you can do this:

new Valify({
    myString: value => {
        if (typeof value !== 'string')
            return 'must be a string';
        if (value.length < 5)
            return 'must be greater than 5 chars';
        if (value.length > 10)
            return 'must be less than 10 chars';
    }
})

Arguments in custom type

• Inside all custom type function are passed 3 arguments:
  • value, current value
  • data, a copy of origin data object
  • be, a library used for several validations. More info on beJS

Example

new Valify({
    color0: 'string',
    color1: (value, data, be) => {
        if (!be.string(value))
            return 'must be a string';
        if (value === data.color0)
            return 'must be different of color0';
    }
})

Locale

You can set locale string in two ways:

1) Globally, using static method setLocale

Valify.setLocale({
    TYPE_FAIL: 'this type has failed'
});

Default strings

|Name|Default| |-|-| |UNKNOWN_TYPE|Unknown type: "{type}"| |TYPE_FAIL|"{path}" expects "{type}" but receives: {dataField}| |TYPE_ARRAY_FAIL|"{path}" expects array of "{type}" but receives: {dataField}| |TYPE_FUNCTION_FAIL|"{path}" receives: {dataField}| |FIELD_REQUIRED|"{path}" is required| |DATA_REQUIRED|Data is required and must be an object| |FIELD_CANNOT_EMPTY|"{path}" cannot be empty| |UNKNOWN_DETECTED|Unknown fields were detected: {unknown}|

2) Local, into field settings

// Define a model
const userModel = new Valify({
    aString: {
        type: 'string',
        locale: {
            TYPE_FAIL: 'this type has failed'
        }
    }
});

• There are only two available properties:
  • TYPE_FAIL
  • TYPE_ARRAY_FAIL
  • FIELD_REQUIRED
  • FIELD_CANNOT_EMPTY

Available types

All types that you can use:

• JavaScript standard
  • arguments
  • array
  • boolean
  • buffer
  • date
  • error
  • float32array
  • float64array
  • function
  • generatorfunction
  • int16array
  • int32array
  • int8array
  • map
  • null
  • number
  • object
  • promise
  • regexp
  • set
  • string
  • symbol
  • uint16array
  • uint32array
  • uint8array
  • uint8clampedarray
  • undefined
  • weakmap
  • weakset
• Extra
  • alpha
  • alphanumeric
  • any
  • datestring
  • email
  • float
  • int
  • ip
  • timestring
  • uuid
  • url

Upgrade to V4

• Breaking changes
  • Changed in locale strings {field} with {path}
  • Changed convert position, now is before all checks
  • Removed validators, use custom types instead

Changelog

You can view the changelog here

License

Valify is open-sourced software licensed under the MIT license

Author

Fabio Ricali