[null,null,["最后更新时间 (UTC):2025-07-27。"],[],[],null,["# Types of CI automation\n\nThe following are some typical forms of automation that you might like to use in\nyour CI system.\n\nBasic jobs\n----------\n\n- **Build:** By building a project from scratch, you make sure that the new\n changes compile correctly and that all libraries and tools are compatible with\n each other.\n\n- **Lint or style checks:** This is an optional but recommended step. When you\n enforce style rules and perform [static analysis](/studio/write/lint), code reviews can be more\n concise and focused.\n\n- **[Local, or host-side tests](/training/testing/local-tests):** They run on the local machine that\n performs the build. On Android this is usually the JVM, so they're fast and\n reliable. They include Robolectric tests as well.\n\nInstrumented tests\n------------------\n\n[Tests that run on emulators or physical devices](/training/testing/instrumented-tests) require some provisioning,\nwaiting for devices to boot or be connected and other operations that add\ncomplexity.\n\nThere are multiple options to run instrumented tests on CI:\n\n- [Gradle Managed Devices](/studio/test/gradle-managed-devices) can be used to define the devices to use (for example \"Pixel 2 emulator on API 27\") and it handles device provisioning.\n- Most CI systems come with a third-party plugin (also called \"action\", \"integration\" or \"step\") to handle Android emulators.\n- Delegate instrumented tests to a device farm such as [Firebase Test Lab](https://firebase.google.com/docs/test-lab). Device farms are used for their high reliability and they can run on emulators or physical devices.\n\nPerformance regression tests\n----------------------------\n\nTo monitor app performance we recommend using the [benchmark libraries](/topic/performance/benchmarking/benchmarking-overview).\nAutomation of performance tests during development requires physical devices to\nensure consistent and realistic test results.\n\nRunning benchmarks can take a long time, especially when you have high coverage\nof code and user journeys that you are benchmarking. Instead of running all\nbenchmarks for every merged feature or commit, consider executing them as part\nof a regularly scheduled maintenance build, such as a nightly build.\n\n### Monitoring performance\n\nYou can monitor performance regressions using [step fitting](https://medium.com/androiddevelopers/fighting-regressions-with-benchmarks-in-ci-6ea9a14b5c71). Step\nfitting defines a rolling window of previous build results which you compare\nagainst the current build. This approach combines several benchmark results into\none regression-specific metric. You can apply step fitting to reduce noise\nduring regression testing.\n\nThis reduces the occurrence of false positives which can occur when benchmark\ntimes are slow for a single build and then normalize again.\n\nTest coverage regression checks\n-------------------------------\n\n[Test coverage](/studio/test/test-in-android-studio#view_test_coverage) is a metric that can help you and your team decide if tests\nsufficiently cover a change. However, it shouldn't be the only indicator. It is\ncommon practice to set up a regression check that fails or shows a warning when\nthe coverage goes down relative to the base branch.\n| **Note:** The coverage generated by an instrumentation test is different from that of a unit test as bigger tests typically make fewer assertions per line of tested code and their goal is different. Consider keeping two separate coverage metrics."]]