How 4399AT Enables a Practical Android CI/CD Automation Pipeline

CI/CD Overview

Continuous Integration (CI) automates the build and verification of code changes. Continuous Delivery (CD) extends CI by automatically preparing artifacts for release, while Continuous Deployment further pushes verified changes to production without manual intervention. The 4399AT pipeline currently implements CI and Continuous Delivery.

Toolchain

Jenkins – orchestrates the pipeline via plugins for email, GitLab, Gradle, Sonar, etc.

Gradle – compiles the Android project according to the local machine and project configuration.

4399AT – Android UI automation testing tool that executes test cases and generates reports.

Sonar – performs static code analysis on the source.

GitLab – hosts the source repository and triggers Jenkins builds through webhooks.

Tomcat and Nginx – act as reverse‑proxy servers; Tomcat serves LAN access, Nginx provides domain‑based access to avoid IP‑range restrictions.

JSON – defines the structure of test case scripts.

Python – drives overall automation, including test execution, report migration, build result evaluation, and extraction of email parameters.

JavaScript – formats and generates the final HTML reports.

Testing Scenario and Workflow

Instead of building on every commit, the team adopts an environment‑named branch strategy. Developers create branches such as test or production. After local self‑testing, they push code to the test branch and send a test email. The email triggers a GitLab webhook, which starts a Jenkins job.

The Jenkins pipeline performs the following steps:

Checkout the code from the test branch.

Run gradle assemble to compile the Android APK.

Execute Sonar analysis to detect code quality issues.

Invoke 4399AT with the JSON test case scripts to run UI automation.

Collect 4399AT test reports and migrate them to a shared location.

Generate a consolidated HTML report using JavaScript.

Send an email containing the build status, Sonar metrics, and links to the test reports.

The overall flow is illustrated in the diagram below:

Execution Results

Sample test reports generated by 4399AT are shown below:

Impact and Benefits

The pipeline resolves installation failures and crashes of test packages that occurred in earlier manual processes. As the test case library matures, the system can perform functional and exhaustive testing, reducing manual effort and allowing the team to focus on high‑risk modules. Although a precise efficiency metric is not available, the overall cycle time has been reduced by approximately 10‑20% compared to the previous workflow, and the setup continues to evolve.