c8df15e97d413d8f35f91e24f69ad8532ba4be78 - third_party/gcloud

commit	c8df15e97d413d8f35f91e24f69ad8532ba4be78	[log] [tgz]
author	Josh Soref <2119212+jsoref@users.noreply.github.com>	Wed Sep 11 09:36:57 2024 -0400
committer	GitHub <noreply@github.com>	Wed Sep 11 15:36:57 2024 +0200
tree	6fd3d38837fa735b93e6b3c02015396a428fc5b9
parent	eee174cdbfaeaef0c977f0b38e26c2d0a85ce4b5 [diff]

tree: 6fd3d38837fa735b93e6b3c02015396a428fc5b9

README.md

The gcloud package provides a high level idiomatic Dart interface to some of the most widely used Google Cloud Platform services.

Google Cloud Platform support package (gcloud)

Currently the following services are supported:

Cloud Datastore
Cloud Storage
Cloud Pub/Sub

Status: Experimental

NOTE: This package is currently experimental and published under the labs.dart.dev pub publisher in order to solicit feedback.

For packages in the labs.dart.dev publisher we generally plan to either graduate the package into a supported publisher (dart.dev, tools.dart.dev) after a period of feedback and iteration, or discontinue the package. These packages have a much higher expected rate of API and breaking changes.

Your feedback is valuable and will help us evolve this package. For general feedback, suggestions, and comments, please file an issue in the bug tracker.

API details

The APIs in this package are all based on the generic generated APIs in the googleapis and googleapis_beta packages.

This means that the authentication model for using the APIs in this package uses the googleapis_auth package.

Note that this package is only intended for being used with the standalone VM in a server or command line application. Don't expect this package to work on the browser or in Flutter.

The code snippets below demonstrating the use of this package all assume that the following imports are present:

import 'dart:io';

import 'package:googleapis_auth/auth_io.dart' as auth;
import 'package:http/http.dart' as http;
import 'package:gcloud/db.dart';
import 'package:gcloud/storage.dart';
import 'package:gcloud/pubsub.dart';
import 'package:gcloud/service_scope.dart' as ss;
import 'package:gcloud/datastore.dart' as datastore;

Getting access to the APIs

The first step in using the APIs is to get an authenticated HTTP client and with that create API class instances for accessing the different APIs. The code below assumes that you have a Google Cloud Project called my-project with credentials for a service account from that project stored in the file my-project.json.

// Read the service account credentials from the file.
var jsonCredentials = new File('my-project.json').readAsStringSync();
var credentials = new auth.ServiceAccountCredentials.fromJson(jsonCredentials);

// Get an HTTP authenticated client using the service account credentials.
var scopes = []
    ..addAll(datastore.Datastore.Scopes)
    ..addAll(Storage.SCOPES)
    ..addAll(PubSub.SCOPES);
var client = await auth.clientViaServiceAccount(credentials, scopes);

// Instantiate objects to access Cloud Datastore, Cloud Storage
// and Cloud Pub/Sub APIs.
var db = new DatastoreDB(
    new datastore.Datastore(client, 's~my-project'));
var storage = new Storage(client, 'my-project');
var pubsub = new PubSub(client, 'my-project');

All the APIs in this package supports the use of ‘service scopes’. Service scopes are described in details below.

ss.fork(() {
  // register the services in the new service scope.
  registerDbService(db);
  registerStorageService(storage);
  registerPubSubService(pubsub);

  // Run application using these services.
});

The services registered with the service scope can now be reached from within all the code running in the same service scope using the below getters.

dbService.
storageService.
pubsubService.

This way it is not necessary to pass the service objects around in your code.

Use with App Engine

The gcloud package is also integrated in the Dart appengine package. This means the gcloud services are available both via the appengine context and service scopes. The authentication required to access the Google Cloud Platform services is handled automatically.

This means that getting to the App Engine Datastore can be through either the App Engine context

var db = context.services.db;

or just using the service scope registration.

var db = dbService;

Cloud Datastore

Google Cloud Datastore provide a NoSQL, schemaless database for storing non-relational data. See the product page https://cloud.google.com/datastore/ for more information.

The Cloud Datastore API provides a mapping of Dart objects to entities stored in the Datastore. The following example shows how to annotate a class to make it possible to store instances of it in the Datastore.

@db.Kind()
class Person extends db.Model {
  @db.StringProperty()
  String name;

  @db.IntProperty()
  int age;
}

The Kind annotation tell that instances of this class can be stored. The class must also inherit from Model. Now to store an object into the Datastore create an instance and use the commit function.

var person = new Person()
    ..name = ''
    ..age = 42;
await db.commit(inserts: [person]);

The function query is used to build a Query object which can be run to perform the query.

var persons = (await db.query<Person>().run()).toList();

To fetch one or multiple existing entities, use lookup.

var key = new Person()
    ..name = 'UniqueName'
    ..age = 42
    ..parentKey = db.emptyKey;
var person = (await db.lookup<Person>([key])).single;
var people = await db.lookup<Person>([key1, key2]);

NOTE: This package include a lower level API provided through the class Datastore on top of which the DatastoreDB API is build. The main reason for this additional API level is to bridge the gap between the different APIs exposed inside App Engine and through the public REST API. We reserve the rights to modify and maybe even remove this additional layer at any time.

Cloud Storage

Google Cloud Storage provide a highly available object store (aka BLOB store). See the product page https://cloud.google.com/storage/ for more information.

In Cloud Storage the objects (BLOBs) are organized in buckets. Each bucket has a name in a global namespace. The following code creates a new bucket named my-bucket and writes the content of the file my-file.txt to the object named my-object.

var bucket = await storage.createBucket('my-bucket');
new File('my-file.txt').openRead().pipe(bucket.write('my-object'));

The following code will read back the object.

bucket.read('my-object').pipe(new File('my-file-copy.txt').openWrite());

Cloud Pub/Sub

Google Cloud Pub/Sub provides many-to-many, asynchronous messaging. See the product page https://cloud.google.com/pubsub/ for more information.

Cloud Pub/Sub uses two concepts for messaging. Topics are used if you want to send messages and subscriptions are used to subscribe to topics and receive the messages. This decouples the producer of a message from the consumer of a message.

The following code creates a topic and sends a simple test message:

var topic = await pubsub.createTopic('my-topic');
await topic.publishString('Hello, world!')

With the following code a subscription is created on the topic and a message is pulled using the subscription. A received message must be acknowledged when the consumer has processed it.

var subscription =
    await pubsub.createSubscription('my-subscription', 'my-topic');
var pullEvent = await subscription.pull();
print(pullEvent.message.asString);
await pullEvent.acknowledge();

It is also possible to receive messages using push events instead of pulling from the subscription. To do this the subscription should be configured as a push subscription with an HTTP endpoint.

await pubsub.createSubscription(
    'my-subscription',
    'my-topic',
    endpoint: Uri.parse('https://server.example.com/push'));

With this subscription all messages will be send to the URL provided in the endpoint argument. The server needs to acknowledge the reception of the message with a 200 OK reply.

Running tests

If you want to run the end-to-end tests, a Google Cloud project is required. When running these tests the following environment variables need to be set:

GCLOUD_E2E_TEST_PROJECT

The value of the environment variable GCLOUD_E2E_TEST_PROJECT is the name of the Google Cloud project to use. Authentication for testing uses Application Default Credentials locally you can provide GOOGLE_APPLICATION_CREDENTIALS or use gcloud auth application-default login.

You will also need to create indexes as follows:

gcloud --project "$GCLOUD_E2E_TEST_PROJECT" datastore indexes create test/index.yaml