Monday Code Challenge: Transform data tables

1_mixes.feature

Background:
    Given the following mixes data
        | 1 | Metal | Heavy metal | Track A |
        |   |       |             | Track B |
        |   |       |             | Track C |
        |   |       | Hair metal  | Track D |
        |   |       |             | Track E |
        |   |       |             | Track F |
        | 2 | Dance | House       | Track G |
        |   |       |             | Track H |
        |   |       |             | Track I |
        |   |       | Trance      | Track J |
        |   |       |             | Track K |
        |   |       |             | Track L |

2_data-output.js

// This is the JS output from cucumber from the above data table

var data = [
    ['1',  'Metal',  'Heavy metal',  'Track A'],
    ['',   '',       '',             'Track B'],
    ['',   '',       '',             'Track C'],
    ['',   '',       'Hair metal',   'Track D'],
    ['',   '',       '',             'Track E'],
    ['',   '',       '',             'Track F'],
    ['2',  'Dance',  'House',        'Track G'],
    ['',   '',       '',             'Track H'],
    ['',   '',       '',             'Track I'],
    ['',   '',       'Trance',       'Track J'],
    ['',   '',       '',             'Track K'],
    ['',   '',       '',             'Track L']
];

3_convert.js

// I want this data structure created. You need to look into generating the ids (_.uniqueId() for e.g.)

var groups = [
  {
    id: 123,                          // generate id
    name: 'Metal',
    category: 'genre',                // based off 1 in the first column
    mixes: [
      {
        id: '001',                    // generate id
        name: 'Heavy Metal',
        tracks: [
          { name: 'Track A' },
          { name: 'Track B' },
          { name: 'Track C' }
        ]
      },
      {
        id: '002',                    // generate id
        name: 'Hair Metal',
        tracks: [
          { name: 'Track D' },
          { name: 'Track E' },
          { name: 'Track F' }
        ]
      }
    ]
  },
  {
    id: 234,                          // generate id
    name: 'Dance',
    category: 'theme',                // based off 2 in the first column
    mixes: [
      {
        id: '003',                    // generate id
        name: 'House',
        tracks: [
          { name: 'Track G' },
          { name: 'Track H' },
          { name: 'Track I' }
        ]
      },
      {
        id: '004',                    // generate id
        name: 'Trance',
        tracks: [
          { name: 'Track J' },
          { name: 'Track K' },
          { name: 'Track L' }
        ]
      }
    ]
  }
];

First stab at a solution:

data.reduce(function (collection, row) {
  var x = _.zipObject(['category', 'group', 'mix', 'track'], row);

  if (x.category) {
    collection.push({
      id: _.uniqueId('group'),
      name: x.group,
      category: x.category === 1 ? 'genre' : 'theme',
      mixes: []
    });
  }

  if (x.mix) {
    _.last(collection).mixes.push({
      id: _.uniqueId('mix'),
      name: x.mix,
      tracks: []
    });
  }

  if (x.track) {
    _.last(_.last(collection).mixes).tracks.push({
      name: x.track
    });
  }

  return collection;
}, []);