Appurify Python Client

The official Python client for the Appurify API.

Installation

pip install appurify-0.5.1.tar.gz

This will install any missing dependencies and add two executable scripts to your bin folder:

$ appurify-client.py -h
$ appurify-tunnel.py -h

Running Tests

appurify-client.py --api-key $API_KEY --api-secret $API_SECRET --team $TEAM \
--app-src $APP-SRC --app-test-type $TEST_TYPE --test-src $TEST_SRC --test-type $TEST_TYPE \
--device-type-id $DEVICE_TYPE_IDS --result-dir $RESULT_DIR

Starting Tunnel

appurify-tunnel.py --api-key $API_KEY --api-secret $API_SECRET

To provide local/private network environment to your tests, they must be started after tunnel has been established.

Parameters

• API_KEY: Used for authentication
• API_SECRET: Used for authentication
• TEAM: Name or ID of team to act for (optional)
• APP_SRC: The path or URL to the app binary (.ipa or .apk)
• TEST_SRC: The path or URL where the test files are located
• TEST_TYPE: Your test framework name e.g. calabash, ios_robot, ocunit, uiautomation. See constants.py for list of supported test types.
• DEVICE_TYPE_IDS: A comma separated list of numbers representing the device type IDs you wish to use for your test
• RESULT_DIR: The directory on your local machine where you want your results to be written.

Jenkins Integration

In Jenkins create a new Execute Shell build step and upload your app using the Python wrapper as pictured below.

Exit codes

To facilitate error reporting, the client will report one of the following error codes on exit:

Code	Meaning
0	Test completed with no exceptions or errors
1	Test completed normally but reported test failures
2	Test was aborted by the user or system
3	Test was aborted by the system because of timeout
4	Test could not be completed because the device could not be activated or reserved
5	Test could not execute because there was an error in the configuration or uploaded files
6	Test could not execute because the server rejected the provided credentials
7	Test could not execute because of other server/remote exception
8	Test could not execute because of an unexpected error in the client
9	Test got a connection error attempting to reach the server
10	The app could not be installed on the device (possibly due to incorrect build)
11	Test could not execute because device type is not found in user pool
12	Test could not execute because app is not built for device type
13	Device doesn't exist in users device pool
14	App is not compabtible with device specified.
15	Test reached timeout for grid session

Contribution

Found a bug or want to add a much needed feature? Go for it and send us the Pull Request!

Release Notes

0.5.1

• Support for teams, using the --team flag.

0.5.0

• Loosen dependency requirements so that future versions of required modules are supported

0.4.9

• User-abort now handled smoothly if run has not been uploaded to server and runID not generated.

0.4.8

• An app or test source of size 0 will not be allowed nor uploaded to server.

0.4.5

• Bug fix of multiple devices trying when downloading results flag passed would fail.

0.4.3

• Client will not allow upload of .ipa to android devices
• Client will not allow upload of .apk to iOS devices
• If device type is not in user pool, client will no allow run to upload to server

0#.4

• Better handling around client/server connection errors (including SSL cert errors)

0.3.4

• Add exit codes

0.2.9

• Handle case where test results may not immediately be ready for download after a test completes.

0.2.8

• Added --version flag to print version and exit

0.2.6

• Users will receive a warning when attempting to upload a web test without specifying the url parameter.
• Support for both --timeout parameter to specify the desired timeout at runtime, or using the os environment variable APPURIFY_API_TIMEOUT. Specify desired timeout in seconds.

0.2.2

• Added ios_sencharobot test type

0.2.1

• network_headers test type no longer requires app source
• Fixed an issue where test results were not properly downloaded despite setting the result-dir parameter.
• Test source is now optional for kif test type
• Improved test status information when polling a running test
• Configuration values are now printed when running a test
• Fixed a bug where ``name``` parameter was not respected for web apps