terraformer

terraformer is a Python wrapper around the Terraform CLI.

Usage

Quick Start

terraformer usage centers around the Workspace object, which can be used to run the typical Terraform commands in that workspace. Just like with the CLI you have to initialize the workspace before you can run plans or applies.

from terraformer import Workspace

workspace = Workspace(path="./")
workspace.init()

results, plan = workspace.plan()

if !results.successful:
  raise Exception(f"Terraform run failed without output: {results.stdout}")

if plan.deletions > 0:
  raise Exception("Deletions not expected from this plan")

results, apply_log = workspace.apply(plan_path=plan.plan_path, auto_approve=True)

if !results.successful:
  print("Terraform Apply was not successful.")

for resource_name, resource_data in apply_log.resources.items():
  print(f"${resource_name}: ${resource['message']}")

for output_name, output_data in apply_log.outputs.items():
  print(f"{output_name}:\n")
  print(yaml.dumps(output_data))
  print("\n\n\n")

The plan step can be skipped if you are auto approving.

from terraformer import Workspace

workspace = Workspace(path="./")
workspace.init()
results, apply_log = workspace.apply(auto_approve=True)

There is also the capability to parse an existing plan JSON and create a plan object with the sensitive and unknown attributes sanitized in the output.

  import json

  plan_from_json = TerraformPlan("", "tests/terraform/plans/sensitive_plan.json", False)
  for address, resource_change in plan_from_json.changes:
    print(f"address: {address}\n")
    print("before:\n")
    print(json.dumps(resource_change.before_sanitized))
    print("\nafter:\n")
    print(json.dumps(resource_change.after_sanitized))

Installation

terraformer can be installed with pip.

pip install terraformer

In addition to terraformer you should have the Terraform binary installed on your system.