| # [Android Automotive](https://source.android.com/devices/automotive) Developer Tools |
| |
| AADevT contains tools for AA device developers to improve their productivity. |
| |
| * Fully unbundle from Android versions, and use the AOSP trunk based |
| development model. |
| |
| * Bridge the app ecosystem for [Android Automotive OS](https://developer.android.com/training/cars) |
| because device developers are the key producers for the core apps. |
| |
| * Leverage [Android app developer tools](https://developer.android.com/studio/intro) |
| that a few million developers use. |
| |
| [TOC] |
| |
| ## Dev Tools |
| This contains tools designed specifically to simplify the AA device development |
| workflows. |
| |
| * clone_proj.sh to clone a git project for the unbundled development workflows |
| in instead of the whole Android repo. |
| |
| ## Change Reports |
| |
| ### change_report.py |
| change_report.py creates a diff statistic CSV file from 2 versions of a codebase. |
| This is useful when the git commit history is somehow not obtainable. What you |
| need is to get 2 versions of a codebase downloaded on your disk first. |
| |
| * You can compare specific folders of concern for a quick result, or when |
| there is a code patch change. |
| * This skips all symlinks & ignores common repository metadata folders, e.g. |
| .git, etc. |
| * It can take a long time & generates a large CSV file for the whole Android |
| codebase & especially if they are many changes. For example: |
| * Android 11 QPR1 vs QPR2 takes more than 8 min. & generates a 5MB CSV file. |
| * Android 10 QPR3 vs Android 11 QPR2 takes more than 11 min. & generates a |
| 95MB CSV file. |
| * To reduce time, you should always remove **out**, the build output folder first. |
| * For example, to compare Android 11 QPR1 vs QPR2 AOSP codebases on your disk. |
| |
| ``` |
| python3 change_report.py --old_dir ~/android/android11-qpr1-release \ |
| --new_dir ~/android/android11-qpr2-release \ |
| --csv_file ~/change_reports/change_report_android11-qpr1-release_android11-qpr2-release.csv |
| ``` |
| |
| * An output example: [change_report-new_vs_old_codebase.csv](dev/resource/change_report-new_vs_old_codebase.csv) |
| is the change report between **dev/resource/old_codebase** and |
| **new_codebase**. |
| * The **states** are: |
| * SAME = 0 |
| * NEW = 1 |
| * REMOVED = 2 |
| * MODIFIED = 3 |
| * INCOMPARABLE = 4 |
| |
| ### sysui_oem_diff.sh |
| sysui_oem_diff.sh generates a summary of code changes between 2 revisions. |
| Which gives you a rough idea of changes on files and Lines of Code. |
| |
| * The report is especailly useful to discuss the pain points on sysui/notif |
| customization with AAOS team. |
| * For example, to generate the change report for Android 11 to 10 QPR3: [sysui_gcar_android10-qpr3-release_android11-release.txt](dev/resource/sysui_gcar_android10-qpr3-release_android11-release.txt) |
| |
| ``` |
| ./sysui_oem_diff.sh ~/Android/android11-release remotes/aosp/android10-qpr3-release remotes/aosp/android11-release > sysui_gcar_android10-qpr3-release_android11-release.txt |
| ``` |
| |
| ## System Performance Tuning |
| AAOS system performance turning is hard. Here are the tools to make it a bit |
| easier for the device developers. |
| |
| * time_to_init_disp.sh to measure an app's [Time the Initial Display](https://developer.android.com/topic/performance/vitals/launch-time#time-initial) |
| |
| ## [Android Virtual Device as a Development Platform](avd/README.md) |