/dookie

Import / export mongodb databases in JSON

Primary LanguageJavaScriptApache License 2.0Apache-2.0

dookie

Dookie lets you write MongoDB test fixtures in JSON or YAML with extra syntactic sugar (extended JSON, variables, imports, inheritance, etc.).

Note: Dookie requires Node >= 4.0.0. Dookie is not tested with nor expected to work with Node 0.x or io.js.

Circle CI

Examples

Dookie can be used either via require('dookie'); in Node.js, or from the command line as an executable. Dookie's fundamental operations are:

  1. Push - optionally clear out a database and insert some data
  2. Pull - write the contents of a database to a file

Push is more interesting, so let's start with that. You can access the push functionality with the require('dookie').push() function, or ./node_modules/.bin/dookie push from the command line.

It can import YAML data with .push()

Suppose you have a YAML file called file.yml that looks like below.

people:
  - _id:
      # MongoDB extended JSON syntax
      $oid: 561d87b8b260cf35147998ca
    name: Axl Rose
  - _id:
      $oid: 561d88f5b260cf35147998cb
    name: Slash

bands:
  - _id: Guns N' Roses
    members:
      - Axl Rose
      - Slash

Dookie can push this file to MongoDB for you.

    co(function*() {
      const fs = require('fs');
      const yaml = require('js-yaml');

      const contents = fs.readFileSync('./example/basic/file.yml');
      const parsed = yaml.safeLoad(contents);

      const mongodbUri = 'mongodb://localhost:27017/test';
      // Insert data into dookie
      // Or, at the command line:
      // `dookie push --db test --file ./example/basic/file.yml`
      yield dookie.push(mongodbUri, parsed);

      // ------------------------
      // Now that you've pushed, you should see the data in MongoDB
      const db = yield mongodb.MongoClient.connect(mongodbUri);
      const collections = (yield db.listCollections().toArray()).
        map(v => v.name).filter(v => !v.startsWith('system.')).sort();
      assert.equal(collections.length, 2);
      assert.equal(collections[0], 'bands');
      assert.equal(collections[1], 'people');

      const people = yield db.collection('people').find().toArray();
      people.forEach((person) => { person._id = person._id.toString() });
      assert.deepEqual(people, [
        { _id: '561d87b8b260cf35147998ca', name: 'Axl Rose' },
        { _id: '561d88f5b260cf35147998cb', name: 'Slash' }
      ]);
      const bands = yield db.collection('bands').find().toArray();
      assert.deepEqual(bands, [
        { _id: `Guns N' Roses`, members: ['Axl Rose', 'Slash'] }
      ]);
    })
  

It can $require external files

Suppose you're a more advanced user and have some collections you want to re-use between data sets. For instance, you may want a common collection of users for your data sets. Dookie provides a $require keyword just for that. Suppose you have a file called parent.yml:

$require: ./child.yml

bands:
  - _id: Guns N' Roses
    members:
      - Axl Rose

This file does a $require on child.yml, which looks like this:

people:
  - _id: Axl Rose

When you push parent.yml, dookie will pull in the 'people' collection from child.yml as well.

    co(function*() {
      const filename = './example/$require/parent.yml';
      const contents = fs.readFileSync(filename);
      const parsed = yaml.safeLoad(contents);

      const mongodbUri = 'mongodb://localhost:27017/test';
      // Insert data into dookie
      // Or, at the command line:
      // `dookie push --db test --file ./example/basic/parent.yml`
      yield dookie.push(mongodbUri, parsed, filename);

      // ------------------------
      // Now that you've pushed, you should see the data in MongoDB
      const db = yield mongodb.MongoClient.connect(mongodbUri);

      const people = yield db.collection('people').find().toArray();
      assert.deepEqual(people, [{ _id: 'Axl Rose' }]);

      const bands = yield db.collection('bands').find().toArray();
      assert.deepEqual(bands, [
        { _id: `Guns N' Roses`, members: ['Axl Rose'] }
      ]);
    })
  

It supports inheritance via $extend

You can also re-use objects using the $extend keyword. Suppose each person in the 'people' collection should have a parent pointer to the band they're a part of. You can save yourself some copy/paste by using $extend:

$gnrMember:
  band: Guns N' Roses

people:
  - $extend: $gnrMember
    _id: Axl Rose
  - _id: Slash
    $extend: $gnrMember
    co(function*() {
      const filename = './example/$extend.yml';
      const contents = fs.readFileSync(filename);
      const parsed = yaml.safeLoad(contents);

      const mongodbUri = 'mongodb://localhost:27017/test';
      // Insert data into dookie
      // Or, at the command line:
      // `dookie push --db test --file ./example/$extend.yml`
      yield dookie.push(mongodbUri, parsed, filename);

      // ------------------------
      // Now that you've pushed, you should see the data in MongoDB
      const db = yield mongodb.MongoClient.connect(mongodbUri);

      const people = yield db.collection('people').find().toArray();
      assert.deepEqual(people, [
        { band: `Guns N' Roses`, _id: 'Axl Rose' },
        { band: `Guns N' Roses`, _id: 'Slash' }
      ]);
    })
  

It can evaluate code with $eval

Dookie also lets you evaluate code in your YAML. The code runs with the current document as the context.

people:
  - _id: 0
    firstName: Axl
    lastName: Rose
    name:
      $eval: this.firstName + ' ' + this.lastName
    co(function*() {
      const filename = './example/$eval.yml';
      const contents = fs.readFileSync(filename);
      const parsed = yaml.safeLoad(contents);

      const mongodbUri = 'mongodb://localhost:27017/test';
      // Insert data into dookie
      // Or, at the command line:
      // `dookie push --db test --file ./example/$eval.yml`
      yield dookie.push(mongodbUri, parsed, filename);

      // ------------------------
      // Now that you've pushed, you should see the data in MongoDB
      const db = yield mongodb.MongoClient.connect(mongodbUri);

      const people = yield db.collection('people').find().toArray();
      assert.deepEqual(people, [
        { _id: 0, firstName: 'Axl', lastName: 'Rose', name: 'Axl Rose' }
      ]);
    })
  

It can pull() data out of MongoDB

The above examples show how dookie can push() data into MongoDB. Dookie can also pull() data out of MongoDB in JSON format. Why not just use mongoexport or mongodump? Mongoexport can only export a single collection, mongodump exports hard-to-read binary data, and neither can be run from Node without .exec(). Dookie lets you transfer whole databases in a human readable format, and assert() on the entire state of your database in tests with ease.

    co(function*() {
      const mongodbUri = 'mongodb://localhost:27017/test';
      // Insert data into dookie
      // Or, at the command line:
      // `dookie pull --db test --file ./output.json`
      const json = yield dookie.pull(mongodbUri);

      assert.deepEqual(Object.keys(json), ['people']);
      assert.deepEqual(json.people, [
        { _id: 0, firstName: 'Axl', lastName: 'Rose', name: 'Axl Rose' }
      ]);
    })