Skip to content

Latest commit

 

History

History
231 lines (157 loc) · 11.5 KB

CONTRIBUTING.md

File metadata and controls

231 lines (157 loc) · 11.5 KB

Contributing to Milvus

Contributions to Milvus are welcome from everyone. We strive to make the contribution process simple and straightforward. Up-to-date information can be found at milvus.io.

The following are a set of guidelines for contributing to Milvus. Following these guidelines makes contributing to this project easy and transparent. These are mostly guidelines, not rules. Use your best judgment, and feel free to propose changes to this document in a pull request.

As for everything else in the project, the contributions to Milvus are governed by our Code of Conduct.

Content

What contributions can you make?

Suitable for Projects Resources
Go developers milvus, milvus-sdk-go
CPP developers milvus, knowhere
Developers interested in other languages pymilvus, milvus-sdk-node, milvus-sdk-java Contributing to PyMilvus
Kubernetes enthusiasts milvus-helm
Tech writers and docs enthusiasts milvus-docs Contributing to milvus docs
Web developers milvus-insight

How can you contribute?

Contributing code

If you encounter a bug, you can

  • (Recommended) File an issue about the bug.
  • Provide clear and concrete ways/scripts to reproduce the bug.
  • Provide possible solutions for the bug.
  • Pull a request to fix the bug.

If you're interested in existing issues, you can

  • (Recommended) Provide answers for issue labeled question.
  • Provide help for issues labeled bug, improvement, and enhancement by
    • (Recommended) Ask questions, reproduce the issue, or provide solutions.
    • Pull a request to fix the issue.

If you require a new feature or major enhancement, you can

  • (Recommended) File an issue about the feature/enhancement with reasons.
  • Provide an MEP for the feature/enhancement.
  • Pull a request to implement the MEP.

If you are a reviewer/approver of Milvus, you can

  • Participate in PR review process.
  • Instruct newcomers in the community to complete the PR process.

If you want to become a contributor of Milvus, submit your pull requests! For those just getting started, see GitHub workflow below.

All submissions will be reviewed as quickly as possible. There will be a reviewer to review the codes, and an approver to review everything aside the codes, see code review for details. If everything is perfect, the reviewer will label /lgtm, and the approver will label /approve. Once the 2 labels are on your PR, and all actions pass, your PR will be merged into base branch automatically by our @sre-ci-robot

GitHub workflow

Generally, we follow the "fork-and-pull" Git workflow.

  1. Fork the repository on GitHub.
  2. Clone your fork to your local machine with git clone git@github.com:<yourname>/milvus.git.
  3. Create a branch with git checkout -b my-topic-branch.
  4. Commit changes to your own branch, then push to GitHub with git push origin my-topic-branch.
  5. Submit a pull request so that we can review your changes.

Remember to sync your forked repository before submitting proposed changes upstream. If you have an existing local repository, please update it before you start, to minimize the chance of merge conflicts.

git remote add upstream git@github.com:milvus-io/milvus.git
git fetch upstream
git checkout upstream/master -b my-topic-branch

General guidelines

Before submitting your pull requests for review, make sure that your changes are consistent with the coding style, and run unit tests to check your code coverage rate.

  • Include unit tests when you contribute new features, as they help to prove that your code works correctly, and also guard against future breaking changes to lower the maintenance cost.
  • Bug fixes also require unit tests, because the presence of bugs usually indicates insufficient test coverage.
  • Keep API compatibility in mind when you change code in Milvus. Reviewers of your pull request will comment on any API compatibility issues.
  • When you contribute a new feature to Milvus, the maintenance burden is (by default) transferred to the Milvus team. This means that the benefit of the contribution must be compared against the cost of maintaining the feature.

Developer Certificate of Origin (DCO)

All contributions to this project must be accompanied by acknowledgment of, and agreement to, the Developer Certificate of Origin. Acknowledgment of and agreement to the Developer Certificate of Origin must be included in the comment section of each contribution and must take the form of Signed-off-by: {{Full Name}} <{{email address}}> (without the {}). Contributions without this acknowledgment will be required to add it before being accepted. If contributors are unable or unwilling to agree to the Developer Certificate of Origin, their contribution will not be included.

Contributors sign-off that they adhere to DCO by adding the following Signed-off-by line to commit messages:

This is my commit message

Signed-off-by: Random J Developer <random@developer.example.org>

Git also has a -s command line option to append this automatically to your commit message:

$ git commit -s -m 'This is my commit message'

Coding Style

Keeping a consistent style for code, code comments, commit messages, and PR descriptions will greatly accelerate your PR review process. We highly recommend you refer to and comply to the following style guides when you put together your pull requests:

Golang coding style

We also use golangci-lint to perform code check. Run the following command before submitting your pull request and make sure there is no issue reported:

$ make static-check

To format code

$ make fmt

C++ coding style

The C++ coding style used in Milvus generally follows Google C++ Style Guide. We made the following changes based on the guide:

  • 4 spaces for indentation
  • Adopt .cpp file extension instead of .cc extension
  • 120-character line length
  • Camel-Cased file names

Install clang-format

$ sudo apt-get install clang-format

Check code style

$ make cppcheck

Run unit test with code coverage

Before submitting your Pull Request, make sure you have run unit test, and your code coverage rate is >= 90%.

Run golang unit tests

You can run all the Golang unit tests using make.

$ make test-go

You can also run unit tests in package level.

# run unit tests in datanode package
$ go test ./internal/datanode -cover
ok  	github.com/milvus-io/milvus/internal/datanode 3.874s	coverage: 88.2% of statements

You can run a sub unit test.

In this case, we are only concerned about the tests with name "SegmentReplica" and sub tests with name "segmentFlushed". When running sub tests, the coverage is not concerned.

$ go test ./internal/datanode -run SegmentReplica/segmentFlushed
ok  	github.com/milvus-io/milvus/internal/datanode 0.019s

Using mockery

It is recommended to use Mockery to generate mock implementations for unit test dependencies.

If your PR changes any interface definition, you shall run following commands to update all mockery implemented type before submitting it:

make generate-mockery

If your PR adds any new interface and related mockery types, please add a new entry under proper Makefile generate-mockery-xxx command.

generate-mockery-xxx: getdeps
    # ...
    # other mockery generation commands
    # use mockery under $(INSTALL_PATH) to unify mockery binary version
    $(INSTALL_PATH)/mockery --name=NewInterface ...

Run C++ unit tests

Install lcov

$ sudo apt-get install lcov

Run unit test and generate code for code coverage check

$ make codecov-cpp

Commits and PRs