Tags:

project

Removing vals
Keeping vals
Keeping Subfields
Re-Assign Fields Making New Fields

Project can be thought of like map in js.
Project shapes a new output in this step of the pipeline. Project can be re-used in a bunch in pipelines.

Basic project syntax is...

db.solarSystem.aggregate([{$project: {...aggExpressionshere...}}])

Remove values

Project can remove fields from being presented after this pipeline step.

{
  fieldName: 0;
}

Keep Values

Project can explicitly keep values to keep after this pipeline step. Once a single filed is identified in this projection field, ALL DESIRED FIELDS must be included in the project stage. Leaving out a field means the field will not appear in the output (access for the '_id' field).

{
  fieldname: 1;
}

an example of keeping 1 field, and even removing the _id field:

db.solarSystem.aggregate([
  {
   $project: {_id:0, name:1}
  }
])

// should return...
{ "name" : "Sun" }
{ "name" : "Mercury" }
{ "name" : "Venus" }
{ "name" : "Earth" }
{ "name" : "Mars" }
{ "name" : "Jupiter" }
{ "name" : "Saturn" }
{ "name" : "Uranus" }
{ "name" : "Neptune" }

NOTE: the _id field STAYS in project unless explicitly removed.

Keep Subfields

With subfields, the key(s) containing the subfields must be a string with dot notation, parentObj.childKey:

db.solarSystem.aggregate([{$project: {_id:0, name:1, 'gravity.value': 1}}])
{ "name" : "Sun", "gravity" : { "value" : 274 } }
{ "name" : "Mercury", "gravity" : { "value" : 3.24 } }
{ "name" : "Venus", "gravity" : { "value" : 8.87 } }
{ "name" : "Earth", "gravity" : { "value" : 9.8 } }
{ "name" : "Mars", "gravity" : { "value" : 3.71 } }
{ "name" : "Jupiter", "gravity" : { "value" : 24.79 } }
{ "name" : "Saturn", "gravity" : { "value" : 10.44 } }
{ "name" : "Uranus", "gravity" : { "value" : 8.87 } }
{ "name" : "Neptune", "gravity" : { "value" : 11.15 } }

Reassigning fields

Here, the sub-field of gravity.value will be re-assigned to value.
The new field gets declared as a key. The value that the project uses to assign to the new key.

Prefix field vals with $

The aggregation expression must prefix the field (& subfield combo if present, like below) with a $.
Here, assign the $gravity.value sub-field from the original doc to a different field in the output called gravity. Notice that the original field is written with the prefix $:

db.solarSystem.aggregate([
  {
    $project: {
      _id:0,
      name:1,
      gravity: '$gravity.value'
    }
  }
])

// will return...
{ "name" : "Sun", "gravity" : 274 }
{ "name" : "Mercury", "gravity" : 3.24 }
{ "name" : "Venus", "gravity" : 8.87 }
{ "name" : "Earth", "gravity" : 9.8 }
{ "name" : "Mars", "gravity" : 3.71 }
{ "name" : "Jupiter", "gravity" : 24.79 }
{ "name" : "Saturn", "gravity" : 10.44 }
{ "name" : "Uranus", "gravity" : 8.87 }
{ "name" : "Neptune", "gravity" : 11.15 }

Creating an entirely new field

The new name of a reassigned field is arbitrary, and does not need to match the original field names at all. Here, the gravity.value sub-field gets reassigned to a new surfaceGravity key/value pair -

db.solarSystem.aggregate([
  {$project: {
    _id:0,
    name:1,
    surfaceGravity: '$gravity.value'
  }
}])

// will return
{ "name" : "Sun", "surfaceGravity" : 274 }
{ "name" : "Mercury", "surfaceGravity" : 3.24 }
{ "name" : "Venus", "surfaceGravity" : 8.87 }
{ "name" : "Earth", "surfaceGravity" : 9.8 }
{ "name" : "Mars", "surfaceGravity" : 3.71 }
{ "name" : "Jupiter", "surfaceGravity" : 24.79 }
{ "name" : "Saturn", "surfaceGravity" : 10.44 }
{ "name" : "Uranus", "surfaceGravity" : 8.87 }
{ "name" : "Neptune", "surfaceGravity" : 11.15 }

Dabbling with value calculations

Figure out how much I weigh on each planet.
With a weight of 140lb, 140 will be multiplied by a new gravityRatio.

the gravityRatio will be calculated by gravity.value divided by 9.8, earths gravitational force calculation
the gravityRatio will then be multiplied by 140
this resulting value will be assigned to myWeight

// vars up front
let EARTH_GRAVITY = 9.8;
let gravityRatio = { $divide: ["$gravity.value",9.8] };
let MY_WEIGHT = 140;
let myWeightOnPlanet = { $multiply: [gravityRatio, MY_WEIGHT] };

// passing 'myWeightOnPlanet' as val of `myWeightHere`
db.solarSystem.aggregate([{
  $project: {
    _id:0,
    planet: "$name",
    myWeightHere: myWeightOnPlanet
  }
}])

// all-in-one
db.solarSystem.aggregate([
  {$project: {
    _id:0,
    name:1,
    myWeight: {
      $multiply: [
        {
          $divide: ["$gravity.value",9.8]
        },
        140
      ]
    }
  }}
])

{ "name" : "Sun", "myWeight" : 3914.285714285714 }
{ "name" : "Mercury", "myWeight" : 46.28571428571429 }
{ "name" : "Venus", "myWeight" : 126.7142857142857 }
{ "name" : "Earth", "myWeight" : 140 }
{ "name" : "Mars", "myWeight" : 53 }
{ "name" : "Jupiter", "myWeight" : 354.1428571428571 }
{ "name" : "Saturn", "myWeight" : 149.14285714285714 }
{ "name" : "Uranus", "myWeight" : 126.7142857142857 }
{ "name" : "Neptune", "myWeight" : 159.28571428571428 }

Pipelines and JS

Pipeline arrays can be saved as vars && inserted into the mongodb query module.

let weightOnPlanetsProjection = [
  {
    $project: {
      _id: 0,
      name: 1,
      myWeight: {
        $multiply: [
          {
            $divide: ['$gravity.value', 9.8],
          },
          140,
        ],
      },
    },
  },
];

// use it something like this...
mongoNodeMod.aggregate(weightOnPlanetsProjection);

Nested objects and arrays

Here, 2 different formats of a user object in a collection, where the stop field content is either an obj or an arr ->

{
  _id: 1,
  user: "1234",
  stop: {
    title: "book1",
    author: "xyz",
    page: 32
  } }
{
  _id: 2,
  user: "7890",
  stop: [
    {
      title: "book2",
      author: "abc",
      page: 5
    }, {
      title: "book3",
      author: "ijk",
      page: 100
    }
  ]
}

A single projection will "work" to project only the book titles ->

db.bookmarks.aggregate([{ $project: { 'stop.title': 1 } }]);
// or
db.bookmarks.aggregate([{ $project: { stop: { title: 1 } } }]);