Thank you for sharing your code with the ADAM project. We appreciate your contribution!
If you're not already on the ADAM developers list, take a minute to join. It would be great if you'd introduce yourself to the group but it's not required. You can just let your code do the talking for you if you like.
You can find us on Freenode IRC in the #adamdev room.
Before you write too much code, check the open issues in the ADAM issue tracker or open pull requests to see if someone else is already working on it. If the issue is a new one, and you've already written the code to address the issue, simply submit a pull request. If you haven't written any code but plan to, go ahead and open a new issue and assign it to yourself to socialize your work. When you submit your PR, please reference this pull request so that it will be automatically closed when the PR is merged.
The issue tracker has a list of "pick me up!" issues that are good tasks to help you get familiar with ADAM; they don't require an understanding of genomics and are fairly limited in scope.
Shoot us a quick email on the mailing list letting us know what you're working on. There will likely be people on the list who can give you tips about where to find relevant source or alert you to other planned changes that might effect your work.
If the work you're proposing makes substantive changes to ADAM, you may be asked to attach a design document to your issue in the issue tracker. This document should provide a high-level explanation of your design, clearly define the goal of the new design and explain the expected effects on performance, APIs, etc. This document is meant to save you time as it allows the team a chance to provide feedback on the proposes changes. It's likely we can help you find a way to achieve your goals with less work. The document also allows the team to prepare for large changes to the code base. We welcome change but also want to ensure that code quality is kept high.
Github provides a nice overview on how to create a pull request.
Some general rules to follow:
- Do your work in a fork of the ADAM repo.
- Create a branch for each feature/bug in ADAM that you're working on. These branches are often called "feature" or "topic" branches.
- Use your feature branch in the pull request. Any changes that you push to your feature branch will automatically be shown in the pull request. If your feature branch is not based off the latest master, you will be asked to rebase it before it is merged. This ensures that the commit history is linear, which makes the commit history easier to read.
- Remember to reference any issues that your pull request fixes in the commit message. This will ensure that the issue is automatically closed when the PR is merged.
- Run the ./scripts/format-source script in order to format the code and ensure correct license headers
$ ./scripts/format-source
- Please alphabetize your imports. We follow the following approach: first, alphabetize by package name (e.g.,
org.apache.spark
should be beforeorg.bigdatagenomics.adam
). Within a project, "lower level" packages should be sorted ahead (e.g.,org.apache.spark.SparkContext
should be beforeorg.apache.spark.rdd.RDD
). Within a single package, sort alphabetically, but put object implicit imports first (e.g., putorg.apache.spark.SparkContext._
beforeorg.apache.spark.Logging
, andorg.apache.spark.Logging
beforeorg.apache.spark.SparkContext
). - Keep your pull requests as small as possible. Large pull requests are hard to review. Try to break up your changes into self-contained and incremental pull requests, if need be, and reference dependent pull requests, e.g. "This pull request builds on request #92. Please review #92 first."
- The first line of commit messages should start by referencing the issue number they fix (i.e., "[ADAM-307]" indicates that this commit fixes ADAM issue #307), followed by a short (<80 character) summary, followed by an empty line and then, optionally, any details that you want to share about the commit.
- Include unit tests with your pull request. We love tests and use Jenkins
to check every pull request and commit. Just look for files in the ADAM repo that end in "*Suite.scala",
e.g. ADAMContextSuite.scala, to see examples of how to write tests. You might also want to glance at the
./scripts/jenkins-test
script for more end-to-end tests.