Thread-safe client functionality for gcloud-python via requests.

pip install --upgrade gcloud_requests

Note that at this time, only gcloud==0.17.0 on Python 2.7 is officially supported.

The library provides new HTTP objects that can be passed in to the gcloud.*.Client constructors (per supported API module). For example, to use the connection class (with proper retrying implemented) for Google Cloud Datastore:

from gcloud import datastore
from gcloud_requests.connection import datastore_http

client = datastore.Client(http=datastore_http)
client.query(kind="EntityKind").fetch()

and for the Google Cloud Storage service:

from gcloud import storage
from gcloud_requests.connection import storage_http

client = storage.Client(http=storage_http, project="my-project")
bucket = client.get_bucket("my-bucket")

Using a custom HTTP object causes gcloud Clients to ignore whatever Credentials objects you pass into them manually. If you need to use a custom set of credentials with gcloud_requests you must instantiate a RequestsProxy object by passing in those credentials and then passing that instance to your client like so:

from gcloud import datastore
from gcloud_requests.requests_connection import DatastoreRequestsProxy

http = DatastoreRequestsProxy(custom_credentials)
client = datastore.Client(http=http)

The gcloud-python library for accessing Google Cloud Platform services like Google Cloud Datastore, Google Cloud Storage, Google BigQuery, and others, relies on the httplib2 library to handle the underlying Protobuf requests. This library (httplib2) is not threadsafe.

Based on notes in gcloud-python#926, gcloud-python#908, and gcloud-python#1214, this library replaces the underlying transport with requests.

It's strongly encouraged that you let Travis run the tests. This is because running the tests requires gcd, the Google Cloud Datastore tools, which also requires a Google Cloud Platform service key...if you're still intrigued, work through the gcd documentation and then simply install pytest and run the tests with py.test.

gcloud_requests was authored at Leadpages. You can find out more about contributors here. We welcome contributions, and we're always looking for more engineering talent!

Please read our contributor's guide.