This repository serves as a central hub for the Frinx Machine backend service APIs, represented in Pydantic format. Its primary objectives are to maintain a historical record of API alterations, enhance the visibility of version-to-version changes, and bolster compatibility maintenance within conductor workers.
To transform REST APIs, we rely on Swagger as our source of truth. We employ the datamodel-codegen tool for this transformation process.
When it comes to transforming GraphQL schemas, we extract the JSON schema directly from the service. To accomplish this, we make use of the graphql-pydantic-converter tool.
In this repository is housed several Python packages. Each of these packages serves as a Pydantic representation of either a REST API or a GraphQL schema, tailored to a specific backend service.
-
Whenever a new version of the backend service is released, it triggers an automatic GitHub action. This action initiates the transformation process, converting the API from either REST or GraphQL format.
-
Upon successful transformation, a GitHub action generates a Pull Request (PR) containing the changes.
-
Subsequently, tests are expected to be executed on this PR to ensure the integrity and functionality of the newly transformed code.
-
After the PR is reviewed and approved, it is merged into the main branch.
-
Once merged, another GitHub action is triggered, responsible for releasing a package with an associated tag to maintain version control. Only changes in RELEASE.md invoke tag creation.
The best course of action would be to refer to your package manager's documentation for comprehensive guidance. However, here are some straightforward examples of usage with Poetry.
# inventory package, example with branch usage
poetry add git+ssh://[email protected]/FRINXio/frinx-services-python-api.git@main#subdirectory=inventory/python
# inventory package, example with tag usage
poetry add git+ssh://[email protected]/FRINXio/[email protected]#subdirectory=inventory/pythonOnce you've installed the apis, you can use them in your Python projects. Import the required modules and use the provided functionality to build requests with input validation.
import requests
from typing import Any
from frinx_api.uniconfig.rest_api import DryrunCommit
from frinx_api.uniconfig.dryrun.manager.dryruncommit import Input
from frinx_api.uniconfig.dryrun.manager.dryruncommit import TargetNodes
def class_to_json(dataclass: Any) -> Any:
return dataclass.json(exclude_none=True, by_alias=True, exclude_defaults=True)
response = requests.request(
url='http://uniconfig:8181/rests' + DryrunCommit.uri,
method=DryrunCommit.method,
data=class_to_json(
DryrunCommit.request(
input=Input(
do_rollback=True,
target_nodes=TargetNodes(
node=['IOS01']
)
)
)
)
)