fs_service_lib 2.0.3

Firestore service library. Allows to create, delete, update documents and collections using service account.

fs_service_lib

A Dart library for interacting with Google Cloud Firestore REST API via googleapis. It simplifies creating, reading, updating, and deleting Firestore documents and collections with built-in JSON mapping, typed value prefixes, and configurable in-memory HTTP response caching.

---

Features

• Firestore REST API Integration: Direct interaction with Cloud Firestore using googleapis and standard Google Auth credentials.
• Collection Queries & Filtering: Powerful collection filtering via queryCollection using FirestoreFilter and FirestoreOperator (isEqualTo, isGreaterThan, arrayContains, etc.).
• Collection Group Queries: Easily query subcollections across all documents by collectionId with sorting and pagination.
• Batch Operations: Atomic batchWrite for multiple update and delete operations in a single HTTP request.
• Batch Read: Fast getDocumentsByIds fetching multiple specific documents by ID in one REST call.
• Aggregations & Existence: Built-in countCollection for document counts and documentExists for fast presence checks.
• Subcollection Discovery: Inspection of document subcollections via getCollectionIds.
• Simplified JSON Mapping: Convert raw JSON maps into Firestore documents and collections seamlessly, including recursive nested collections.
• Typed Value Prefixes: Easily handle non-standard JSON types (timestamps, geopoints, document references, and byte arrays) using simple string prefixes:
  • datetime:// for UTC ISO-8601 Timestamps
  • location:// for GeoPoints (latitude/longitude)
  • reference:// for Document References
  • bytes:// for Base64 Binary Data
• Metadata Management: Automatic injection and extraction of document metadata (e.g. $name, $createTime, $updateTime, $collections, $documents). Custom metadata prefixes are supported.
• Smart In-Memory HTTP Caching:
  • Transparent HTTP caching layer via CachingHttpClient.
  • Flexible cache invalidation strategies (smart, all, none).
  • Automatic invalidation on write/delete operations while safely bypassing read-only query endpoints.
  • Comprehensive cache statistics tracking (hits, misses, stores, invalidations, hitRatioPercentage).

---

Getting Started

Add fs_service_lib to your pubspec.yaml dependencies:

dependencies:
  fs_service_lib: ^2.0.0

Import the library in your Dart code:

import 'package:fs_service_lib/fs_service_lib.dart';

---

Authentication & Environment Setup

fs_service_lib uses clientViaApplicationDefaultCredentials from googleapis_auth to authenticate requests against Google Cloud APIs (Firestore).

To run the application successfully, your environment must provide Google Application Default Credentials (ADC) through one of the following methods:

1. Service Account Key File (Environment Variable)

Set the GOOGLE_APPLICATION_CREDENTIALS environment variable to the absolute path of your GCP Service Account JSON key file:

export GOOGLE_APPLICATION_CREDENTIALS="/path/to/service-account-key.json"

Note

Ensure the service account has appropriate IAM roles (e.g., Cloud Datastore User or Firebase Admin) for accessing your Firestore database.

2. Google Cloud Managed Infrastructure

When running on Google Cloud Platform (GCP) services such as Cloud Run, Google Kubernetes Engine (GKE), Compute Engine (GCE), App Engine, or Cloud Functions, credentials are acquired automatically via the GCP Metadata Server. No GOOGLE_APPLICATION_CREDENTIALS environment variable is required if the runtime service account has the necessary permissions.

3. Local Development with gcloud CLI

For local development without downloading a JSON key file, authenticate using the Google Cloud SDK CLI:

gcloud auth application-default login

This creates local user credentials (typically saved at ~/.config/gcloud/application_default_credentials.json on Linux/macOS or %APPDATA%\gcloud\application_default_credentials.json on Windows) which googleapis_auth discovers automatically.

---

Usage

1. Initialization

Initialize FsServiceLib with your Firebase/Google Cloud projectId and optional parameters:

final fsService = FsServiceLib();

Future<void> main() async {
  await fsService.init(
    projectId: 'your-firebase-project-id',
    databaseId: '(default)', // Default database ID
    cachingClient: CachingHttpClientMemory(
      ttl: const Duration(minutes: 30),
      cacheCheckPeriod: const Duration(minutes: 5),
      invalidationStrategy: CacheInvalidationStrategy.smart,
      onCacheHit: (url, stats) =>
          print('CACHE HIT: $url (${stats.hitRatioPercentage})'),
    ),
  );
}

2. Document & Collection Operations

Get a Single Document

final doc = await fsService.db.getDocument(
  documentPath: '/users/user-123',
);

print(doc);
// Output includes document fields + metadata:
// {
//   'name': 'John Doe',
//   '$name': 'user-123',
//   '$createTime': '2026-01-01T12:00:00.000Z',
//   '$updateTime': '2026-01-02T12:00:00.000Z'
// }

Check Document Existence

final exists = await fsService.db.documentExists(
  documentPath: '/users/user-123',
);
print('Exists: $exists');

Get Multiple Documents by IDs

final docs = await fsService.db.getDocumentsByIds(
  collectionPath: '/users',
  documentIds: ['user-123', 'user-456'],
);
print(docs);

Fetch a Collection

final collection = await fsService.db.getCollection(
  collectionPath: '/users',
  pageSize: 20,
  orderBy: 'name asc',
);

print(collection['$documents']); // List of document JSON objects

Query Collection with Filters

final activeUsers = await fsService.db.queryCollection(
  collectionPath: '/users',
  filters: [
    const FirestoreFilter(
      field: 'age',
      op: FirestoreOperator.isGreaterThanOrEqualTo,
      value: 18,
    ),
    const FirestoreFilter(
      field: 'status',
      op: FirestoreOperator.isEqualTo,
      value: 'active',
    ),
  ],
  limit: 10,
  orderBy: 'age',
  descending: false,
);

Count Documents in Collection

final totalCount = await fsService.db.countCollection(
  collectionPath: '/users',
  filters: [
    const FirestoreFilter(
      field: 'status',
      op: FirestoreOperator.isEqualTo,
      value: 'active',
    ),
  ],
);
print('Active users count: $totalCount');

Query a Collection Group

final posts = await fsService.db.getCollectionGroup(
  collectionId: 'posts',
  limit: 20,
  orderBy: 'dateTime',
  descending: true,
);

print(posts); // List of document JSON maps across all `posts` subcollections

Get Subcollection Names

final collectionIds = await fsService.db.getCollectionIds(
  documentPath: '/users/user-123',
);
print('Subcollections: $collectionIds');

Add a Document

await fsService.db.addDocument(
  collectionPath: '/users',
  json: {
    r'$name': 'user-123', // Optional custom document ID
    'name': 'Alice Smith',
    'age': 30,
    'location': 'location://37.7749/-122.4194',
    'createdAt': 'datetime://2026-08-05T18:00:00.000Z',
    'avatar': 'bytes://AAECAwQFBg==',
  },
);

Update a Document

await fsService.db.updateDocument(
  documentPath: '/users/user-123',
  json: {
    'age': 31,
  },
);

Atomic Batch Write (Update & Delete)

await fsService.db.batchWrite(
  writes: [
    const UpdateWrite(
      documentPath: '/users/user-123',
      json: {'status': 'active'},
    ),
    const DeleteWrite(
      documentPath: '/users/user-old',
    ),
  ],
);

Delete a Document or Collection

// Delete a specific document
await fsService.db.deleteDocument(
  documentPath: '/users/user-123',
);

// Delete an entire collection
await fsService.db.deleteCollection(
  collectionPath: '/tempCollection',
);

---

Data Type Prefixes

fs_service_lib uses prefixed strings to map complex Firestore data types in standard JSON maps:

Data Type	Prefix Syntax	Example
Timestamp	datetime://<ISO-8601>	'datetime://2026-08-05T18:00:00Z'
GeoPoint	location://<lat>/<lon>	'location://37.7749/-122.4194'
Reference	reference://<path>	'reference://projects/my-proj/databases/(default)/documents/users/123'
Bytes	bytes://<base64>	'bytes://AAECAwQFBg=='

Custom prefixes can be configured when initializing FsServiceLib.

---

Metadata Keys

Document metadata fields are prefixed with $ by default:

• $name: Document ID or path.
• $createTime: Creation timestamp (UTC ISO-8601 string).
• $updateTime: Last modification timestamp (UTC ISO-8601 string).
• $documents: List of documents within a collection response.
• $collections: List of nested subcollections within a document JSON response.

---

HTTP Caching & Invalidation

fs_service_lib supports transparent HTTP caching via the abstract CachingHttpClient interface to reduce Firestore REST API read operations and bandwidth costs.

Configuring CachingHttpClientMemory

To enable in-memory caching, pass a configured instance of CachingHttpClientMemory when initializing FsServiceLib (or FirestoreApiProviderImpl):

final cachingClient = CachingHttpClientMemory(
  ttl: const Duration(minutes: 15),
  maxEntrySizeBytes: 512 * 1024, // 512 KB per response limit
  maxSizeBytes: 50 * 1024 * 1024, // 50 MB total memory limit
  cacheCheckPeriod: const Duration(minutes: 5),
  invalidationStrategy: CacheInvalidationStrategy.smart,
  onCacheHit: (url, stats) => print('CACHE HIT: $url'),
);

await fsService.init(
  projectId: 'your-firebase-project-id',
  cachingClient: cachingClient,
);

Configuring CachingHttpClientHive (Persistent Disk Caching)

For persistent disk caching across application restarts using hive_ce, initialize Hive and pass an open Box to CachingHttpClientHive:

import 'package:hive_ce/hive.dart';

// Initialize Hive and open a cache box
Hive.init('/path/to/cache_dir');
final box = await Hive.openBox('http_cache');

final cachingClient = CachingHttpClientHive(
  box: box,
  ttl: const Duration(minutes: 30),
  maxEntrySizeBytes: 2 * 1024 * 1024, // 2 MB per response limit
  maxSizeBytes: 50 * 1024 * 1024,      // 50 MB total disk cache limit
  cacheCheckPeriod: const Duration(minutes: 10),
  invalidationStrategy: CacheInvalidationStrategy.smart,
  onCacheHit: (url, stats) => print('DISK CACHE HIT: $url'),
);

await fsService.init(
  projectId: 'your-firebase-project-id',
  cachingClient: cachingClient,
);

Caution

Cloud Run & Serverless Environments: In Cloud Run or Cloud Functions, local files (e.g. in /tmp) use an in-memory filesystem (tmpfs). Writing files to disk consumes container RAM! Unless an external volume is mounted, set maxSizeBytes appropriately to avoid Out-Of-Memory (OOM) errors.

If cachingClient is omitted or passed as null, caching is completely disabled and all HTTP requests go directly to the Firestore REST API.

Direct Cache Control & Memory Inspection

Since CachingHttpClient is managed directly, stats, memory usage, and cache operations are accessed on the CachingHttpClient instance:

// Inspect cache performance stats & current memory consumption
print(cachingClient.stats.hitRatioPercentage);
print('Current cache memory: ${cachingClient.currentSizeBytes} bytes');

// Manually flush all cached entries
cachingClient.clearCache();

Cache Invalidation Strategies

• smart (default): Clears the mutated document's cached URL and any parent collection cached responses upon successful mutating requests (PATCH, DELETE, non-query POST).
• all: Flushes the entire cache whenever a write or delete operation occurs.
• none: Bypasses auto-invalidation (requires manual cache management).

---

Clean Up

Remember to dispose of resources when shutting down:

await fsService.dispose();