Replace with

Description:

Replaces fields in the existing document with the fields in a given expression. The expression can either be a literal map or an expression that evaluates to a map. If a given expression does not evaluate to a map, this stage returns an error.

Syntax:

Mode: {full_replace | merge_overwrite_existing | merge_keep_existing}

replace_with(map: Expr, mode: Mode)

Mode Behaviour:

  • full_replace: Replaces the entire document with the result of the given expression, omitting fields that don't appear in it.
  • merge_overwrite_existing: Merges the expression with the existing document, overwriting field values in the document with values in the expression
  • merge_keep_existing: Merges the expression with the existing document, only adding field values when one does not exist in the document.

Examples:

Create a cities collection with the following documents:

Node.js

await db.collection('cities').doc('SF').set({name: 'San Francisco', population: 800000, location: {country: 'USA', state: 'California'}});
await db.collection('cities').doc('TO').set({name: 'Toronto', population:  3000000, province: 'ON', location: {country: 'Canada', province: 'Ontario'}});
await db.collection('cities').doc('NY').set({name: 'New York', location: {country: 'USA', state: 'New York'}});
await db.collection('cities').cov('AT').set({name: 'Atlantis', population: null});

Using the full_replace mode to get a mutated version of the document

Extract the nested location field, discarding all other data:

Node.js
  const names = await db.pipeline()
    .collection("/cities")
    .replace_with(Field.of("location"), "full_replace")
    .execute();
Java
Pipeline.Snapshot names =
    firestore.pipeline().collection("cities").replaceWith(field("location")).execute().get();
PipelineSnippets.java

Which produces the following documents:

{country: 'USA', state: 'California'},
{country: 'Canada', province: 'Ontario'},
{country: 'USA', state: 'New York'},
{}

Using the merge_overwrite_existing mode to set fields

Set the population field of all documents to 0, overwriting existing values:

Node.js
  const censoredResults = await db.pipeline()
    .collection("/cities")
    .replace_with(map("population", 0), "merge_overwrite_existing")
    .execute();

Which produces the following documents:

{name: 'San Francisco', population: 0, location: {country: 'USA', state: 'California'}},
{name: 'Toronto', population: 0, province: 'ON', location: {country: 'Canada', province: 'Ontario'}},
{name: 'New York', population: 0, location: {country: 'USA', state: 'New York'}},
{name: 'Atlantis', population: 0}

Using the merge_keep_existing mode to add a default value

Setting a default value for location when it doesn't appear in a document:

Node.js
  const defaultedResults = await db.pipeline()
    .collection("/cities")
    .replace_with(map("location", "unknown"), "merge_keep_existing")
    .execute();

Which produces the following documents:

{name: 'San Francisco', population: 800000, location: {country: 'USA', state: 'California'}},
{name: 'Toronto', province: 'ON', population: 3000000, location: {country: 'Canada', province: 'Ontario'}},
{name: 'New York', location: {country: 'USA', state: 'New York'}},
{name: 'Atlantis', population: null, location: "unknown"}