This is also the place where the files are actually being served. This has the advantage, that the files could be modified by the users. Changes Fixed the broken code blocks in the CI part of the documentation.
Since it is now the plan that all relevant generated reports etc would be able to be viewed remotely over a web server, it would make sense to also generate the necessary html for this in the specific classes. Having both of these seperately now is more consistent, becauses not every test run has to be part of a build, but every build contains a test run.
The implementation for this new process is in the module ufotest. This description is then rendered to the test report as well The test reports and the build reports now also save as JSON files. The very same stylesheet is also being used for the test report page. These values are needed to display them in the web pages of the CI server. Refactored the testing procedure: Deprecated the old implementations of TestRunner, TestReport and TestMetadata and replaced them with new implementations.
Added the new class TestContext. The way the testing process is now implemented internally functions much the same way as the build system. A new context manager has to be created for each new test run. This context is then first passed to the test runner, which actually executes the tests and writes the results back to the context. The jinja templates for the CI web server now use template inheritance, which was used to add a default header to each web page, which can be used to navigate between the most important pages.
Added a test result class which allows to add dicts as a result. Test case which simply requests a frame from the camera and adds the image to the test report. Test case which requests frame and calculates simple statistics for it.
The mock test case thus became a way of testing test results classes. Fixed the combined test result not having an html conversion.
Both the hostname and the port now have to be defined in the config file. I realized that the html files for the reports are being created in separate instances of the application, which are not the running server and thus do not know the hostname or the port.
This has lead to errors in the creation of the absolute url links within the html templates. Now there is a build Queue. When a build is triggered, it is added to this queue, which is essentially just a json file with a list. Changes Fixed: The stylesheet for the build report html template was not linked to correctly. Fixed: There was an issue where the program would crash if two build would be triggered at the same hour and minute of the day since they would attempt to create folders with the same name.
Added the build commit and the second of to the folder name format to fix. Fixed the bug that a build would not save the test reports properly. Fixed the build report url within the report email. Documentation Updated the section about continuous integration. Added the option to skip the clone and flash steps for testing purposes. Furhtermore the method does not return a build report anymore, it has to be seperately constructed.
Previously it would always use the default option defined within the config file. Attempting to release the build lock when it is not locked not raises the appropriate exception.
Attempting to construct a test report from a test context without an actual test run being executed before now raises the appropriate exceptions The build lock is now uses the correct folder which can be changed at runtime. The build process now properly fails when the flash command fails. BuildWorker process now does not crash when a KeyboardInterrupt is received.
BuildWorker process does not crash when the build process fails and also not when the email sending fails.
If test suite is currently processing, one cannot access the test archive list in the web interface presumably because the test folder already exists but no reports have been saved to it. The best course for fixing this would be to prevent a crash for an empty test folder alltogether.
Build and test reports are now properly sorted by start date within the web interface FIXED: The test are not executed in the order in which they are specified in the config file FIXED: Not necessarily a bug with this code but with the way the camera works: I feel like there is a problem with how the decoding of the images works.
I feel like all the images accumulate within the internal buffer of the camera and they are ALL being read out with each call to ipedecode? Templating of the actual report files is done at compile time and not during the actual web request. This means that essentially all test reports become unusable when changing the hostname or generally anything within the config which is relevant for the test report.
Add information about the test suite to the builds list page. Parameter for test cases encoded in their names. Could do it by colon separation or smth. Project details Project links Homepage.
Download files Download the file for your platform. Files for ufotest, version 1. Close Hashes for ufotest File type Wheel. Python version py2. Upload date Mar 19, Hashes View. File type Source. Web icon An illustration of a computer application window Wayback Machine Texts icon An illustration of an open book.
Books Video icon An illustration of two cells of a film strip. Video Audio icon An illustration of an audio speaker. Audio Software icon An illustration of a 3. Software Images icon An illustration of two photographs.
Images Donate icon An illustration of a heart shape Donate Ellipses icon An illustration of text ellipses. A collection of declassified UFO files from all countries. Media Type Media Type. Year Year. Collection Collection. Creator Creator. Language Language. It started in , and it was the third study of its kind the first two were projects Sign and Grudge A termination order was given for the study in December , and all activity under its auspices ceased in January
0コメント