Avoid duplicate data selection from arrays, while using hasMany : 'schemaName' inside another object schema

[swatjoshi]

Hi I have been facing problem due to the duplicate data generation in the array, which is used inside another schema. When I say array usage inside another schema, I mean using hasMany : 'schemaName' for populating multiple values of schemaName I am attaching my JS file and related screen-shot to make you understand better Thanks

var mocker = require('mocker-data-generator').default;
  var {ObjectID} = require('mongodb');
  var user = {
    email : {
        faker : 'internet.email'
    },
    org : {
        values : ['deloitte', 'tcs', 'infosys', 'wipro']
    },
    name : {
        faker : 'name.findName'
    },
    _id : {
        function : function () {
            return new ObjectID();
        }
    }
} 
var finalSchema = {
    //I want two unique users to be inserted inside my finalSchema
    userDetails : {
        hasMany : 'users',
        amount : 2 
    },
    date : {
        faker : 'date.past'
    }
}
mocker()
.schema('users', user, {uniqueField :'_id', min : 10, max : 10})
.schema('finalObject', finalSchema, 2)
.build()
.then(function (data) {
    console.log(JSON.stringify(data.finalObject, undefined, 2));
});

issue.zip

[danibram]

First of all thank you for helping to improve mocker. Try the example below in the playground, I think is that you need, the problem is like uniqueField is made only for fields that are array of strings, not for array of objects:

Also I m going to implement unique option in hasMany to make more clear.

For example uniqueField it will work with uniqueField:'org' in the case of user model Thank you, let me if it works!

var user = {
    email : {
        faker : 'internet.email'
    },
    org : {
        values : ['deloitte', 'tcs', 'infosys', 'wipro']
    },
    name : {
        faker : 'name.findName'
    },
    _id : {
        function : function () {
            return Math.floor(Math.random() * 100);
        }
    }
}
var finalSchema = {
userDetails : {
    hasMany : 'users',
        amount : 2
    },
    date : {
        faker : 'date.past'
    }
}

mocker()
    .schema('users', user, 10)
    .schema('finalObject', finalSchema, 2)

[swatjoshi]

Hi Thank you for your reply !! The code you asked me to try in playground, is working the way I want but the problem is that I have requirement that my data should have mongoose id only which is 24 character string with hexa-decimal values For example "_id": "5a1d0ec1ba89971c146af1a1" which is a string and Math.random() may not work for this

[danibram]

yes, but in my playgroud i dont want to install external libs, so i substitiute for a native one. Change for your ObjectId and its done

[swatjoshi]

Hello Again !! I replaced ObjectId with the Math.random(). But it created the duplicates again. Even I tried with Math.random() and they are generated again. I am attaching the screenshots for both Math.random() and ObjectId outputs var mocker = require('mocker-data-generator').default; var {ObjectID} = require('mongodb'); var user = { email : { faker : 'internet.email' }, org : { values : ['deloitte', 'tcs', 'infosys', 'wipro'] }, name : { faker : 'name.findName' }, _id : { function : function () { //return Math.floor(Math.random() * 100); return new ObjectID(); } } } var finalSchema = { //I want two unique users to be inserted inside my finalSchema userDetails : { hasMany : 'users', amount : 2 }, date : { faker : 'date.past' } } mocker() .schema('users', user, 10) .schema('finalObject', finalSchema, 2) .build() .then(function (data) { console.log(JSON.stringify(data.finalObject, undefined, 2)); });

[danibram]

Yes, as i said, hasMany was not be implemented to get unique values, so right now i recomend you to increase the number of users. And im going to develop that feature!Thanks for your contribution!

[swatjoshi]

Oh sure..I will work on your suggestion. Thank you for your guidance !!

[bcl-lcb]

Is the hasOne guarantee get a unique result? I want to create two table that has One to One relationship.

[danibram]

Thanks for the comments. Unfortunately no right now hasMany is based on hasOne, and hasOne take a random element

hasOne(cfg) {
        let db = this.DB
        let i = Math.floor(db[cfg.hasOne].length * Math.random())
        let entity = db[cfg.hasOne][i]

        if (cfg.get) {
            return eval('entity.' + cfg.get)
        } else {
            return entity
        }
    }

    hasMany(cfg) {
        /* ... */
        return Array.from(new Array(amount)).map(() => this.hasOne(newCfg))
    }
}

Rigth now you need to increase the number of the related entity, but, yes, this is not the correct way. Im developing that feature, this issues helps me a lot to know what features need to be released, so if you feel miss something dont be shy, open issues. So thank you again, and stay tunned! I will close this in the next version!

[bcl-lcb]

OK. Thank you! Looking forward about this. It will be awesome if there are ways to associate tables with One to One, One to Many and Many to Many relationship.

[coffenbacher]

Thanks for looking into this! The library looks really good. What I am hoping for to adopt it is some way to generate Many-to-Many relationships that are unique.

For example, imagine a social network with a Users table that is N:N with itself, using a Friendships table as a through table. How can we generate unique Friendships? I believe this is doable via the ask for a HasOne unique constraint.

[danibram]

Hi guys, I just released the version 2.6.6 that has unique option in hasMany, so I will close that. The feature that @CraigLee17 is talking maybe it includes a rework of the core so I will open another issue about that. Thanks for using mocker-data-generator!