Use the Inheritance Pattern

On this page

About this Task

Steps
Learn More

Use the inheritance pattern when your documents are mostly similar and you want to keep them in the same collection so they can be read together. The inheritance pattern uses a parent entity with common fields to group child entities that have variable forms. The child entities can have unique fields, but are closely related to one another due to their common fields.

About this Task

In this example, a book store uses the inheritance pattern to store different types of media. A book parent entity stores common fields like title and author, and multiple child entities inherit from the book entity. For example audio books, printed books, and ebooks have common fields, and also have unique fields specific to the media type.

The inheritance pattern stores these slightly different entities in the same collection, which improves performance for queries that need to access all books, regardless of type.

Steps

Insert the sample data

db.books.insertMany( [
   {
      product_type: "ebook",
      title: "Practical MongoDB Aggregations",
      author: "Paul Done",
      rating: 4.8,
      genres: [ "programming" ],
      pages: 338,
      download_url: "<url>"
   },
   {
      product_type: "audiobook",
      title: "Practical MongoDB Aggregations",
      author: "Paul Done",
      rating: 4.6,
      genres: [ "programming" ],
      narrators: [ "Paul Done" ],
      duration: {
         hours: 21,
         minutes: 8
      },
      time_by_chapter: [
         {
            chapter: 1,
            start: "00:00:00",
            end: "01:00:00"
         },
         {
            chapter: 2,
            start: "01:00:00",
            end: "01:55:00"
         }
      ]
   },
   {
      product_type: "physical_book",
      title: "Practical MongoDB Aggregations",
      author: "Paul Done",
      rating: 4.9,
      genres: [ "programming" ],
      pages: 338,
      stock: 12,
      delivery_time: 2
   }
] )

The preceding documents share some common fields, and have unique fields depending on the product_type. For example:

ebook documents have a download_url field.
audiobook documents have a time_by_chapter field.
physical_book documents have a delivery_time field.

Query all documents

Even though the documents in the books collection have different shapes, you can return all documents with a single query:

db.books.find()

Output:

[
   {
      _id: ObjectId('66eb4160ef006be6eda8e2ee'),
      product_type: 'ebook',
      title: 'Practical MongoDB Aggregations',
      author: 'Paul Done',
      rating: 4.8,
      genres: [ 'programming' ],
      pages: 338,
      download_url: '<url>'
   },
   {
      _id: ObjectId('66eb4160ef006be6eda8e2ef'),
      product_type: 'audiobook',
      title: 'Practical MongoDB Aggregations',
      author: 'Paul Done',
      rating: 4.6,
      genres: [ 'programming' ],
      narrators: [ 'Paul Done' ],
      duration: { hours: 21, minutes: 8 },
      time_by_chapter: [
         { chapter: 1, start: '00:00:00', end: '01:00:00' },
         { chapter: 2, start: '01:00:00', end: '01:55:00' }
      ]
   },
   {
      _id: ObjectId('66eb4160ef006be6eda8e2f0'),
      product_type: 'physical_book',
      title: 'Practical MongoDB Aggregations',
      author: 'Paul Done',
      rating: 4.9,
      genres: [ 'programming' ],
      pages: 338,
      stock: 132,
      delivery_time: 2
   }
]

Query unique fields

The inheritance pattern does not require additional logic to query on fields that are specific to a particular media type. For example, the following query returns books that have a duration longer than 20 hours, which only applies to the product type audio_book:

db.books.find(
   {
      "duration.hours": { $gt: 20 }
   }
)

Output:

[
   {
      _id: ObjectId('66eb4160ef006be6eda8e2ef'),
      product_type: 'audiobook',
      title: 'Practical MongoDB Aggregations',
      author: 'Paul Done',
      rating: 4.6,
      genres: [ 'programming' ],
      narrators: [ 'Paul Done' ],
      duration: { hours: 21, minutes: 8 },
      time_by_chapter: [
         { chapter: 1, start: '00:00:00', end: '01:00:00' },
         { chapter: 2, start: '01:00:00', end: '01:55:00' }
      ]
   }
]

Learn More

Back

Polymorphic Pattern

Versioning