PyBuildkite

A Python library and client for the Buildkite API.

Usage

To get the package, execute:

pip install pybuildkite

Then set up an instance of the Buildkite object, set you access token, and make any available requests.

from pybuildkite.buildkite import Buildkite, BuildState

buildkite = Buildkite()
buildkite.set_access_token('YOUR_API_ACCESS_TOKEN_HERE')

# Get all info about particular org
org = buildkite.organizations().get_org('my-org')

# Get all running and scheduled builds for a particular pipeline
builds = buildkite.builds().list_all_for_pipeline('my-org', 'my-pipeline', states=[BuildState.RUNNING, BuildState.SCHEDULED])

# Create a build
buildkite.builds().create_build('my-org', 'my-pipeline', 'COMMITSHA', 'master', 
clean_checkout=True, message="My First Build!")

Per-Page Configuration

The per_page parameter controls the number of items returned per API request. This can be useful for organizations with large datasets that experience timeouts.

# Default behavior (100 items per page)
buildkite = Buildkite()

# Custom per_page for smaller responses (useful for large organizations)
buildkite = Buildkite(per_page=25)

Pagination

Buildkite offers pagination for endpoints that return a lot of data. By default this wrapper returns 100 objects per page. However, any request that may contain more than that offers a pagination option.

When with_pagination=True, we return a response object with properties that may have next_page, last_page, previous_page, or first_page depending on what page you're on.

builds_response = buildkite.builds().list_all(page=1, with_pagination=True)

# Keep looping until next_page is not populated
while builds_response.next_page:
    builds_response = buildkite.builds().list_all(page=builds_response.next_page, with_pagination=True)

Artifacts

Artifacts can be downloaded as binary data. The following example loads the artifact into memory as Python bytes and then writes them to disc:

artifacts = buildkite.artifacts()
artifact = artifacts.download_artifact("org_slug", "pipe_slug", "build_no", 123, "artifact")
with open('artifact.bin', 'b') as f:
  f.write(artifact)

Large artifacts should be streamed as chunks of bytes to limit the memory consumption:

stream = artifacts.download_artifact("org_slug", "pipe_slug", "build_no", 123, "artifact", as_stream=True)
with open('artifact.bin', 'b') as f:
  for chunk in stream:
    f.write(chunk)

A unicode text artifact can be turned into a string easily:

text = str(artifact)

License

This library is distributed under the BSD-style license found in the LICENSE file.