ODRepair

scripts to run ODRepair

Running ODRepair

The experiments directory contains all the necessary files to run the whole experiments. All the user need to provide is a file contains the corresponding test.

The victim file contains four column like the following example:

ktuukkan/marine-api,af0003847db9ba822f67d4f1dceb8de3fe63250a,.,net.sf.marineapi.ais.event.AbstractAISMessageListenerTest.testBasicListenerWithUnexpectedMessage

The input file that our experiments use is /experiments/data/victim.csv. This file contains all the 327 tests in our experiments. We also provide a file named data/test.csv which contains only 2 tests for simple use.

bash runAll.sh <path_to_input_file>

After running the shell script, all the results file, files that contain generated tests, and files that contain patches are automatically outputted.

files generated for reference

• repos: contain all the necessaey tools with their correponding branch.
• output: store the result file that contains the polluter, time, polluted field (output/overall_results.csv); output/result.csv is the file that only contains the results for one victim to be fed to the next step.
• allFields: store the fields that exist in the passing order/failing order.
• diffFields: store the fields whose values are different between passing order and failing order.
• gen_tests: store the generated tests through Randoop.
• resetters: store the potential reset-methods found by Randoop.
• patches: store the patches that Patcher generates.