Extensions Terraform

Self-managed Google Cloud Functions that replicate Firebase Extensions functionality, deployed and managed entirely through Terraform.

This approach gives you full control over the infrastructure -- you own the Cloud Function, the IAM bindings, and the configuration, rather than relying on the Firebase Extensions framework.

Available Extensions

firestore-translate-text

A Cloud Function that listens for writes to a Firestore collection and automatically translates text fields into one or more target languages.

Supported translation providers:

• Cloud Translation API (default) -- Google's production translation service
• Gemini via Vertex AI -- uses Genkit with the Vertex AI plugin
• Gemini via Google AI -- uses Genkit with the Google AI plugin (requires an API key)

Key features:

• Translates on document create and update, cleans up on delete
• Supports both single string fields and structured object fields
• Per-document language overrides via a configurable field
• Skips redundant translations when the input hasn't changed
• Publishes Eventarc events for downstream integrations
• Configurable scaling (min/max instances, memory, timeout)

Prerequisites

• Terraform >= 1.0
• Google Cloud SDK (gcloud)
• A Google Cloud project with billing enabled
• A Firestore database in the target project

Quick Start

1. Clone the repository:

   git clone <repo-url>
   cd extensions-terraform

2. Navigate to the Terraform directory for the extension you want to deploy:

   cd firestore-translate-text/terraform

3. Create a terraform.tfvars file from the example:

   cp terraform.tfvars.example terraform.tfvars

   Edit it with your project ID and any other overrides. See variables.tf for all available options.

4. Initialize and apply:

   terraform init
   terraform plan
   terraform apply

   Terraform will enable the required APIs, create a dedicated service account with least-privilege IAM roles, upload the function source, and deploy the Cloud Function with a Firestore event trigger.

5. Test it out by writing a document to the configured collection:

   gcloud firestore documents create \
     --collection=translations \
     --document-id=hello \
     --data='{"input": "Hello, world!"}'

   After a few seconds, the document will be updated with a translated field containing translations for each configured language.

Configuration Reference

Each extension's Terraform variables are documented in its variables.tf. Here are the key variables for firestore-translate-text:

Variable	Default	Description
project_id	(required)	GCP project ID
region	us-central1	GCP region for the function
collection_path	translations	Firestore collection to watch
input_field_name	input	Document field containing text to translate
output_field_name	translated	Document field where translations are written
languages	en,es,de,fr	Comma-separated ISO 639-1 language codes
translation_provider	translate	One of: translate, gemini-googleai, gemini-vertexai
gemini_model	gemini-2.5-flash	Model name when using a Gemini provider
function_memory_mb	256	Memory allocation in MB
function_timeout_seconds	60	Function timeout
function_min_instances	0	Minimum instances (set > 0 to avoid cold starts)
function_max_instances	10	Maximum instances

Project Structure

extensions-terraform/
  firestore-translate-text/
    function/          # Cloud Function source (TypeScript)
      src/
        index.ts       # Entry point and Firestore trigger
        config.ts      # Environment variable configuration
        translate/     # Translation logic (strategy pattern)
        logs/          # Structured logging
      package.json
      tsconfig.json
    terraform/         # Terraform configuration
      main.tf          # Resources (function, IAM, APIs, storage)
      variables.tf     # Input variables
      outputs.tf       # Output values
      providers.tf     # Provider configuration

Working on the Function Code

If you want to modify the Cloud Function source:

cd firestore-translate-text/function
npm install
npm run build

The build compiles TypeScript from src/ to lib/. When you run terraform apply, the source is zipped and uploaded to a GCS bucket. The object name includes a content hash, so any code change triggers a redeployment.

License

Apache 2.0