Skip to content

Latest commit

 

History

History
353 lines (244 loc) · 14.2 KB

CONTRIBUTING.md

File metadata and controls

353 lines (244 loc) · 14.2 KB

Contributing

Thanks for helping us to make ksqlDB even better!

If you have any questions about how to contribute, either create a GH issue or ask your question in the #ksqldb channel in our public Confluent Community Slack (account registration is free and self-service).

Developing ksqlDB

About the Apache Maven wrapper

Development versions of ksqlDB use version ranges for dependencies on other Confluent projects, and due to a bug in Apache Maven, you may find that both Maven and your IDE download hundreds or thousands of pom files while resolving dependencies. They get cached, so it will be most apparent on fresh forks or switching to branches that you haven't used in a while.

Until we are able to get a fix in and released officially, we need to work around the bug. We have added a Maven wrapper script and configured it to download a patched version of Maven. You can use this wrapper simply by substituting ./mvnw instead of mvn for every Maven invocation. The patch is experimental, so if it causes you some trouble, you can switch back to the official release just by using mvn again. Please let us know if you do have any trouble with the wrapper.

You can also configure IntelliJ IDEA to use the patched version of Maven. First, get the path to the patched Maven home:

$ ./mvnw -v
...
Maven home: /home/john/.m2/wrapper/dists/apache-maven-3.8.1.2-bin/1npbma9t0n1k5b22fpopvupbmn/apache-maven-3.8.1.2
...

Then, update your IDEA Maven home in Settings > Build, Execution, Deployment > Build Tools > Maven and set the value of Maven home path to the same directory listed as the "Maven home" above. In the above example, it is /home/john/.m2/wrapper/dists/apache-maven-3.8.1.2-bin/1npbma9t0n1k5b22fpopvupbmn/apache-maven-3.8.1.2, but it may be different on your machine.

Finally, you may need to restart the IDE to get it to reload the Maven binary.

There is likely a similar option available for other IDEs.

Building and running ksqlDB locally

To build and run ksqlDB locally, run the following commands:

$ ./mvnw clean package -DskipTests
$ ./bin/ksql-server-start -daemon config/ksql-server.properties
$ ./bin/ksql

This will start the ksqlDB server in the background and the ksqlDB CLI in the foreground. Check the logs folder for the log files that the server writes including any errors.

If you would rather have the ksqlDB server logs spool to the console, then drop the -daemon switch, and start the CLI in a second console.

Building a released version

The source for standalone ksqlDB versions is tagged with the pattern vN.NN.N-ksqldb. For example, the tag for the 0.18.0 standalone release is v0.18.0-ksqldb.

If you wish to build a released version, run the following commands, replacing 0.18.0 with the desired version:

$ git fetch origin --tags
$ git checkout v0.18.0-ksqldb
$ ./mvnw --settings maven-settings.xml clean package -DskipTests

Note that the --settings maven-settings.xml argument is necessary to ensure all dependencies can be pulled respective to the target version.

Running in an IDE

You should configure your IDE to use the Maven wrapper (see "About the Apache Maven wrapper", above).

You can create a run configuration in your IDE for the main class: io.confluent.ksql.rest.server.KsqlServerMain, using the classpath of the ksqldb-rest-app module, and specifying the path to a config file as the program argument. There is a basic config available in config/ksql-server.properties.

NOTE: At least for IDEA (and maybe for other IDEs), building the project doesn't automatically run the generate-sources Maven phase, so our ANTLR and Avro classes will be missing (unless you happen to have built with Maven lately). You can generate them by running:

$ ./mvnw --projects ksqldb-parser,ksqldb-version-metrics-client generate-sources

If you want to use the CLI with your server, you can either build it ahead of time using the package command above, or you can compile and run the class from a terminal using maven:

$ ./mvnw compile exec:java --projects ksqldb-cli -Dexec.mainClass="io.confluent.ksql.Ksql"

Testing changes locally

To build and test changes locally, run the following commands:

$ ./mvnw verify -T 1.5C

This example showcases the new parallel build feature of Maven 3. If it causes problems for you, you can change the parallelism argument, or drop the -T ... option altogether.

Testing docker image

See comments at the top of the docker compose file in the root of the project for instructions on how to build and run docker images.

How to Contribute

Reporting Bugs and Issues

Report bugs and issues by creating a new GitHub issue. Prior to creating an issue, please search through existing issues so that you are not creating duplicate ones.

Guidelines for Contributing Code, Examples, Documentation

Any change to the public API of ksqlDB, especially the SQL syntax, requires a KsqlDB improvement proposal (KLIP) to be raised and approved. See the KLIP Readme for more info.

Code changes are submitted via a pull request (PR). When submitting a PR use the following guidelines:

  • Follow the style guide below
  • Add/update documentation appropriately for the change you are making. For more information, see the docs readme.
  • Non-trivial changes should include unit tests covering the new functionality and potentially function tests.
  • All SQL syntax changes and enhancements should come with appropriate function tests.
  • Bug fixes should include a unit test or integration test potentially function tests proving the issue is fixed.
  • Try to keep pull requests short and submit separate ones for unrelated features.
  • Keep formatting changes in separate commits to make code reviews easier and distinguish them from actual code changes.

Code Style

The project uses GoogleStyle code formatting. You can install this code style into your IDE to make things more automatic:

The project also uses immutable types where ever possible. If adding new type(s), ideally make them immutable.

Static code analysis

The project build runs checkstyle and findbugs as part of the build.

You can set up IntelliJ for CheckStyle. First install the CheckStyle IDEA plugin, then:

IntelliJ → Preferences → Tools → CheckStyle

In top left corner select CheckStyle version 8.44 or newer (older versions fail to parse the provided XML)

- Add a new configurations file using the '+' button:
   Description: Confluent Checks
   URL: https://raw.githubusercontent.com/confluentinc/common/master/build-tools/src/main/resources/checkstyle/checkstyle.xml
   Ignore invalid certs: true

- (Optional) Make the new configuration active.

- Highlight the newly added 'Confluent Checks' and click the edit button (pencil icon).

- Set properties to match the `checkstyle/checkstyle.properties` file in the repo.

'Confluent Checks' will now be available in the CheckStyle tool window in the IDE and will auto-highlight issues in the code editor.

Commit messages

The project uses [Conventional Commits][https://www.conventionalcommits.org/en/v1.0.0-beta.4/] for commit messages in order to aid in automatic generation of changelogs. As described in the Conventional Commmits specification, commit messages should be of the form:

<type>[optional scope]: <description>

[optional body]

[optional footer]

where the type is one of the following, and determines whether or not the commit will appear in the auto-generated changelog for releases. Commit types that do appear in changelogs are:

  • "fix": for bug fixes
  • "feat": for new features
  • "perf": for performance enhancements
  • "revert": for reverting other changes

Commit types that do not appear in changelogs include:

  • "docs": for docs-only changes
  • "test": for test-only changes
  • "refactor": for refactors
  • "style": for stylistic-only changes
  • "build": for build-related changes
  • "chore": for automated changes, as well as other changes not relevant for users. This can include changes that don't fall cleanly into any of the other categories above such as PRs that contain progress towards an incomplete feature.

The (optional) scope is a noun describing the section of the codebase affected by the change. Examples that could make sense for ksqlDB include "parser", "analyzer", "rest server", "testing tool", "cli", "processing log", and "metrics", to name a few.

The optional body and footer are for specifying additional information, such as linking to issues fixed by the commit or drawing attention to breaking changes.

Breaking changes

A commit is a "breaking change" if users should expect different behavior from an existing workflow as a result of the change. Examples of breaking changes include deprecation of existing configs or APIs, changing default behavior of existing configs or query semantics, or the renaming of exposed JMX metrics. Breaking changes must be called out in commit messages, PR descriptions, and upgrade notes:

  • Commit messages for breaking changes must include a line (in the optional body or footer) starting with "BREAKING CHANGE: " followed by an explanation of what the breaking change was. For example,

    feat: inherit num partitions and replicas from source topic in CSAS/CTAS
    
    BREAKING CHANGE: `CREATE * AS SELECT` queries now create sink topics with partition
    and replica count equal to those of the source, rather than using values from the properties
    `ksql.sink.partitions` and `ksql.sink.replicas`. These properties are now deprecated.
    
  • The breaking change should similarly be called out in the PR description. This description will be copied into the body of the final commit merged into the repo, and picked up by our automatic changelog generation accordingly.

  • [Upgrade notes][https://github.com/confluentinc/ksql/blob/master/docs/installation/upgrading.rst] should also be updated as part of the same PR.

Commitlint

This project has [commitlint][https://github.com/conventional-changelog/commitlint] configured to ensure that commit messages are of the expected format. To enable commitlint, simply run npm install from the root directory of the ksqlDB repo (after [installing npm][https://www.npmjs.com/get-npm].) Once enabled, commitlint will reject commits with improperly formatted commit messages.

GitHub Workflow

  1. Fork the confluentinc/ksql repository into your GitHub account: https://github.com/confluentinc/ksql/fork.

  2. Clone your fork of the GitHub repository, replacing <username> with your GitHub username.

    Use ssh (recommended):

    git clone [email protected]:<username>/ksql.git

    Or https:

    git clone https://github.com/<username>/ksql.git
  3. Add a remote to keep up with upstream changes.

    git remote add upstream https://github.com/confluentinc/ksql.git

    If you already have a copy, fetch upstream changes.

    git fetch upstream
  4. Create a feature branch to work in.

    git checkout -b feature-xxx remotes/upstream/master
  5. Work in your feature branch.

    git commit -a
  6. Periodically rebase your changes

    git pull --rebase
  7. When done, combine ("squash") related commits into a single one

    git rebase -i upstream/master

    This will open your editor and allow you to re-order commits and merge them:

    • Re-order the lines to change commit order (to the extent possible without creating conflicts)
    • Prefix commits using s (squash) or f (fixup) to merge extraneous commits.
  8. Submit a pull-request

    git push origin feature-xxx

    Go to your fork main page

    https://github.com/<username>/ksql

    If you recently pushed your changes GitHub will automatically pop up a Compare & pull request button for any branches you recently pushed to. If you click that button it will automatically offer you to submit your pull-request to the confluentinc/ksql repository.

    • Give your pull-request a meaningful title that conforms to the Conventional Commits specification as described above for commit messages. You'll know your title is properly formatted once the Semantic Pull Request GitHub check transitions from a status of "pending" to "passed".
    • In the description, explain your changes and the problem they are solving. Be sure to also call out any breaking changes as described above.
  9. Addressing code review comments

    Repeat steps 5. through 7. to address any code review comments and rebase your changes if necessary.

    Push your updated changes to update the pull request

    git push origin [--force] feature-xxx

    --force may be necessary to overwrite your existing pull request in case your commit history was changed when performing the rebase.

    Note: Be careful when using --force since you may lose data if you are not careful.

    git push origin --force feature-xxx

Backporting Commits

There might be times when a certain commit needs to be backported to a previous ksqlDB release (either community edition or confluent edition). In these situations, cherry-pick individual commits to the previous branches (do not squash multiple commits and move them as one because our changelog generation tool does not handle squashed commits properly).