As of wily 15.10, some autopilot tests are broken in trunk. As such autopilot failures can be excused until they are fixed in trunk.
The Doc Viewer app follows a test driven development where autopilot tests are run before every merge into trunk. If you are submitting your bugfix/patch to the Doc Viewer app, please follow the following steps below to ensure that all tests pass before proposing a merge request.
If you are looking for more info about Autopilot or writing AP tests for the doc viewer app, here are some useful links to help you:
For help and options on running tests, see:
Install the following autopilot packages required to run the tests,
$ sudo apt-get install python3-autopilot libautopilot-qt ubuntu-ui-toolkit-autopilot python3-autopilot-vis
Using terminal:
-
Branch the Doc Viewer app code, for example,
$ bzr branch lp:ubuntu-docviewer-app
-
Navigate to the tests/autopilot directory.
$ cd ubuntu-docviewer-app/tests/autopilot
-
run all tests.
$ autopilot3 run -vv ubuntu_docviewer_app
-
to list all tests:
$ autopilot3 list ubuntu_docviewer_app
To run only one test (for instance: ubuntu_docviewer_app.tests.test_docviewer.TestMainWindow.test_open_pdf_file)
$ autopilot3 run -vv ubuntu_docviewer_app.tests.test_docviewer.TestMainWindow.test_open_pdf_file
-
Debugging tests using autopilot vis
$ autopilot3 launch -i Qt qmlscene app/ubuntu-docviewer-app.qml
$ autopilot3 vis
Refer this tutorial to run tests on Ubuntu SDK:
Using autopkg:
-
Branch the Doc Viewer app code, for example,
$ bzr branch lp:ubuntu-docviewer-app
-
Navigate to the source directory.
$ cd ubuntu-docviewer-app
-
Build a click package
$ click-buddy .
-
Run the tests on device (assumes only one click package in the directory)
$ adt-run . *.click --- ssh -s adb -- -p