The aim of this project is to explore the boundaries of software test automation in a continuous delivery pipeline by implementing a re-ordering algorithm to optimally sort the test cases such that tests with a higher likely hood of failing are executed first, thus enabling faster feedback for failed tests to developers.

To test it:

• Clone repository

  cd jenkins-test-optimizer

• TO generate testset: Run test_generator.py

  $ ./test_generator.py

  $ ./test_generator.py 200

• The generated tests are stored under 'datasets/' and name of the file containing generated tests:

  $ datasets/testdataset.csv

• To generate multi user datasets from single testdata set use the multiuser_datasets.py:

$ ./multiuser_datasets.py

$ ./multiuser_datasets.py datasets/testdataset_010242.csv 7

• To Re-order test sets using algorithm script: Run test_reordering_algorithm.py

  $ ./test_reordering_algorithm.py datasets/testdataset.csv

• To Re-order multiple testsets use script reorder_multiplesets.sh

  $ ./reorder_multiplesets.sh

  $ # this script works now, but is a WIP

• To compute a sample impact value single pair

  $ ./computations.py

  $ ./computations.py datasets/testdataset_010242.csv datasets/testdataset_010242.csv_sorted.csv

• For multiple sets: WIP

  $ ./computations_multiplesets.sh