This document is an extension to CONTRIBUTING and provides more details about the coding guidelines and requirements.
- Code must be well structured:
- packages must have a limited responsibility (different concerns can go to different packages),
- types must be easy to compose,
- think about maintainbility and testability.
- "Depend upon abstractions, [not] concretions".
- Try to limit the number of methods you are exposing. It's easier to expose something later than to hide it.
- Take advantage of
internal
package concept. - Follow agreed-upon design patterns and naming conventions.
- publicly-exposed functions are named logically, have forward-thinking arguments and return types.
- Avoid global variables and global configurators.
- Favor composable and extensible designs.
- Minimize code duplication.
- Limit third-party dependencies.
Performance:
- Avoid unnecessary operations or memory allocations.
Security:
- Pay proper attention to exploits involving:
- gas usage
- transaction verification and signatures
- malleability
- code must be always deterministic
- Thread safety. If some functionality is not thread-safe, or uses something that is not thread-safe, then clearly indicate the risk on each level.
Make sure your code is well tested:
- Provide unit tests for every unit of your code if possible. Unit tests are expected to comprise 70%-80% of your tests.
- Describe the test scenarios you are implementing for integration tests.
- Create integration tests for queries and msgs.
- Use both test cases and property / fuzzy testing. We use the rapid Go library for property-based and fuzzy testing.
- Do not decrease code test coverage. Explain in a PR if test coverage is decreased.
We expect tests to use require
or assert
rather than t.Skip
or t.Fail
,
unless there is a reason to do otherwise.
When testing a function under a variety of different inputs, we prefer to use
table driven tests.
Table driven test error messages should follow the following format
<desc>, tc #<index>, i #<index>
.
<desc>
is an optional short description of whats failing, tc
is the
index within the test case table that is failing, and i
is when there
is a loop, exactly which iteration of the loop failed.
The idea is you should be able to see the
error message and figure out exactly what failed.
Here is an example check:
<some table>
for tcIndex, tc := range cases {
<some code>
resp, err := doSomething()
require.NoError(err)
require.Equal(t, tc.expected, resp, "should correctly perform X")
We are forming a QA team that will support the core Cosmos SDK team and collaborators by:
- Improving the Cosmos SDK QA Processes
- Improving automation in QA and testing
- Defining high-quality metrics
- Maintaining and improving testing frameworks (unit tests, integration tests, and functional tests)
- Defining test scenarios.
- Verifying user experience and defining a high quality.
- We want to have acceptance tests! Document and list acceptance lists that are implemented and identify acceptance tests that are still missing.
- Acceptance tests should be specified in
acceptance-tests
directory as Markdown files.
- Supporting other teams with testing frameworks, automation, and User Experience testing.
- Testing chain upgrades for every new breaking change.
- Defining automated tests that assure data integrity after an update.
Desired outcomes:
- QA team works with Development Team.
- QA is happening in parallel with Core Cosmos SDK development.
- Releases are more predictable.
- QA reports. Goal is to guide with new tasks and be one of the QA measures.
As a developer, you must help the QA team by providing instructions for User Experience (UX) and functional testing.