Skip to content

Latest commit

 

History

History
170 lines (118 loc) · 7.36 KB

File metadata and controls

170 lines (118 loc) · 7.36 KB

DEVELOPER.md

Versioning

This library follows Semantic Versioning.

Processes

Conventional Commit messages

This repository uses tool Release Please to create GitHub and PyPi releases. It does so by parsing your git history, looking for Conventional Commit messages, and creating release PRs.

Learn more by reading How should I write my commits?

Modifying existing classes

When modifying any existing classes, ensure the required libraries are installed in "editable mode" using the command pip install -e . from your project's root directory.

This allows your code to dynamically reflect any changes you make to the library, enabling efficient local testing of your new methods.

Dependencies

The core dependencies needed for the working of this library are mentioned in requirements.txt, but it's version ranges along with the testing dependencies are set in pyproject.toml.

Testing

Formatting

This repo has a lint checker test which is run on every PR as set up in .github/workflows/lint.yml. You can run the formatter locally by following these steps.

  1. Download all test dependencies

    pip install .[test]
  2. Run the black code formatter

    black .
  3. Run isort to sort all imports

    isort .
  4. Run mypy type checker

    mypy .

Run tests locally

  1. Set environment variables for INSTANCE_ID, CLUSTER_ID, DATABASE_ID, DB_USER, DB_PASSWORD, OMNI_HOST, OMNI_USER, OMNI_PASSWORD, IAM_ACCOUNT.

  2. Run pytest to automatically run all tests:

    pytest

Notes:

  • Tests run against public and private IP addresses. Tests for private IP can not be run locally due to VPC restrictions. There is no current way to prevent these tests from running. These tests will time out.
  • Tests use both IAM and built-in authentication.
    • Learn how to set up a built-in databases user at Manage AlloyDB user roles.
    • Local tests will run against your gcloud credentials. Use gcloud to login with your personal account or a service account. This account will be used to run IAM tests. Learn how to set up access to the database at Manage IAM authentication. The "IAM_ACCOUNT" environment variable is also used to test authentication to override the local account. A personal account or a service account can be used for this test.
    • You may need to grant access to the public schema for your new database user: GRANT ALL ON SCHEMA public TO myaccount@example.com;

AlloyDB Omni Testing

The ScaNN index is an AlloyDB Omni preview and is not available on Cloud AlloyDB. To test for the ScaNN index integration, an AlloyDB Omni instance is set up and run on a GCE VM instance. The Omni instance is listening on input traffic on the private IP address of the VM. Integration tests related to the ScaNN index are only run against this Omni instance, while all other index tests are run against the Cloud AlloyDB testing instance.

For more information, refer to the instruction on AlloyDB Omni setup.

CI Platform Setup

Cloud Build is used to run tests against Google Cloud resources in test project: llama-index-alloydb-testing. Each test has a corresponding Cloud Build trigger, see all triggers. These tests are registered as required tests in .github/sync-repo-settings.yaml.

Trigger Setup

Cloud Build triggers (for Python versions 3.10 to 3.13) were created with the following specs:

name: pg-integration-test-pr-py310
description: Run integration tests on PR for Python 3.10
filename: integration.cloudbuild.yaml
github:
  name: llamaindex-alloydb-pg-python
  owner: googleapis
  pullRequest:
    branch: .*
    commentControl: COMMENTS_ENABLED_FOR_EXTERNAL_CONTRIBUTORS_ONLY
ignoredFiles:
  - docs/**
  - .kokoro/**
  - .github/**
  - "*.md"
substitutions:
  _DATABASE_ID: <ADD_VALUE>
  _INSTANCE_ID: <ADD_VALUE>
  _REGION: us-central1
  _VERSION: "3.10"

Use gcloud builds triggers import --source=trigger.yaml to create triggers via the command line

Project Setup

  1. Create an AlloyDB cluster, instance, and database.
  2. Create a private pool for Cloud Build, in order to connect to the VPC network to run tests against a private IP. Learn more at Run builds in a private pool
  3. Setup Cloud Build triggers (above). Note: Triggers are created in same region as private pool.

Run tests with Cloud Build

  • Run integration test:

    gcloud builds submit --config integration.cloudbuild.yaml --region us-central1 --substitutions=_INSTANCE_ID=$INSTANCE_ID,_CLUSTER_ID=$CLUSTER_ID,_DATABASE_ID=$DATABASE_ID,_REGION=$REGION

Trigger

To run Cloud Build tests on GitHub from external contributors, ie RenovateBot, comment: /gcbrun.

Code Coverage

Please make sure your code is fully tested. The Cloud Build integration tests are run with the pytest-cov code coverage plugin. They fail for PRs with a code coverage less than the threshold specified in .coveragerc. If your file is inside the main module and should be ignored by code coverage check, add it to the omit section of .coveragerc.

Check for code coverage report in any Cloud Build integration test log. Here is a breakdown of the report:

  • Stmts: lines of executable code (statements).
  • Miss: number of lines not covered by tests.
  • Branch: branches of executable code (e.g an if-else clause may count as 1 statement but 2 branches; test for both conditions to have both branches covered).
  • BrPart: number of branches not covered by tests.
  • Cover: average coverage of files.
  • Missing: lines that are not covered by tests.

Documentation

LlamaIndex Integration Docs

Google hosts documentation on LlamaIndex's site for individual integration pages: Vector Stores, Document Store, Index Store, Reader, and Chat Store.

Currently, manual PRs are made to the LlamaIndex GitHub repo.

API Reference

Build the documentation

API docs are templated in the docs/ directory.

To test locally, run: nox -s docs

The nox session, docs, is used to create HTML to publish to googleapis.dev The nox session, docfx, is used to create YAML to publish to CGC.

Publish the documentation

The kokoro docs pipeline runs when a new release is created. See .kokoro/ for the release pipeline.