Add ADB (Android Debug Bridge) Support

[ep1cman]

This pull request add support for ADB (Android Debug Bridge) devices in labgrid.

• It supports USB (remote and local) and TCP adb devices.
• It adds two new client commands
  • adb for interfacing with the device by the adb cli
  • scrcpy for remote screen/touch/keyboard/clipboard/etc. sharing

Checklist

• Documentation for the feature
• Tests for the feature

• The arguments and description in doc/configuration.rst have been updated

• Add a section on how to use the feature to doc/usage.rst

• Add a section on how to use the feature to doc/development.rst

• PR has been tested

• Man pages have been regenerated

[codecov]

Codecov Report

❌ Patch coverage is 31.63265% with 134 lines in your changes missing coverage. Please review.
✅ Project coverage is 45.6%. Comparing base (ec5ec29) to head (29fe2a5).
✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
labgrid/remote/client.py	6.6%	56 Missing ⚠️
labgrid/remote/exporter.py	0.0%	45 Missing ⚠️
labgrid/driver/adb.py	54.1%	33 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff            @@
##           master   #1564     +/-   ##
========================================
- Coverage    45.8%   45.6%   -0.2%     
========================================
  Files         182     184      +2     
  Lines       14720   14913    +193     
========================================
+ Hits         6743    6803     +60     
- Misses       7977    8110    +133     

Flag	Coverage Δ
3.10	45.6% <31.6%> (-0.2%)	⬇️
3.11	45.6% <31.6%> (-0.2%)	⬇️
3.12	45.6% <31.6%> (-0.2%)	⬇️
3.13	45.5% <31.6%> (-0.2%)	⬇️
3.14	45.5% <31.6%> (-0.2%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[Emantor]

Some minor comments in regard to subcommand handling and attribute naming.

[Emantor]

Is this always the same port? How will this interact with multiple devices connected over ADB to the same host?

[ep1cman]

Good point, I had not tested multiple simultaneous screen sharing sessions. I have added some code to find a free port on the exporter and use that. It is based on the "get_free_port" code within labgrid that gets a free port on the client machine.

Is this worth moving into a more general location? Maybe within SSHConnection?

[Emantor]

get_free_port should already be in util.helper, so you should be able to import it directly.

[ep1cman]

That returns a free port on the machine running the labgrid client, I require a free port on the exporter. The way scrcpy works is that it runs a server on the android device and creates a port forward between it and the ADB host (the exporter in this case)

[Emantor]

You might be able to handle that within the remote resource for the exporter, the SerialPortExport does something similar to choose a free port for ser2net.

[ep1cman]

There is an additional complication to the scrcpy command.

I already use get_free_port in the ADBExport to get a port for the ADB server, which is started by the ADBExport.

The scrcpy server is not setup at this time, it is started by the scrcpy client when it connects to the device. It first connects via ADB, transfers over the binaries, starts the server, then uses ADB to create a port forward between the device and the ADB host (the exporter).

I am unable to know at the time of starting the exporter what ports will be available at the time scrcpy is launched, which could be weeks later.

[alextercete]

Any updates on this? It would be great to get this merged.

[ep1cman]

@Emantor I believe there was one remaining comment thread, I provided an explanation for why I did things that way, awaiting your feedback. I have update the PR to resolve merge conflicts with the latest master branch.

[Emantor]

It looks like we are missing DCOs for the additional commits, you can either squash them down into the original commit and give credit there or ask the committers to add the DCO 👍

[ep1cman]

It looks like we are missing DCOs for the additional commits, you can either squash them down into the original commit and give credit there or ask the committers to add the DCO 👍

I have squash the commits down and added the others as co-authors. Only outstanding issue now is the use of processwrapper, I left some questions above.