smart-test-runner

Test selection for code changes works by mapping files from git diff output to predefined test mappings, identifying the minimal affected subset across E2E, contract, unit, integration, and shell tests to execute.

Git diff based test mapping is a process that reads changed files from git diff output and maps them to predefined test suites, identifying the minimal subset of E2E, contract, unit, integration, and shell tests needed to verify those specific changes.

Yes, automated test selection supports mapping changed files to relevant test suites across E2E, contract, unit, integration, and shell tests, providing deterministic commands to execute the minimal affected subset.

Setting up predefined test mappings involves defining deterministic relationships between your source files and corresponding test suites, which the test runner then uses alongside git diff outputs to select and execute the correct minimal tests.

A limitation of using git diff for CI test selection is that it relies entirely on predefined test mappings and git diff outputs; if mappings are incomplete or changes affect shared dependencies unpredictably, the minimal selected subset might miss critical tests.

Running a subset of tests after modifying code saves verification time while maintaining confidence, as the deterministic mapping ensures only the minimal test suites directly affected by the changed files are selected and executed.