PR# 19450 AutoTest extracted test issue
Problem Report Summary
Environment: Windows 10
Synopsis: AutoTest extracted test issue
I have followed all steps in the tutorial on this page to create and execute an extracted test: https://www.eiffel.org/doc/eiffelstudio/Using_extracted_tests Unfortunately the result is not as expected. Instead of getting the expected result saying that the postcondition of feature "withdraw" is violated, I get this result: "test_withdraw (TEST_BANK_ACCOUNT_002): FAIL (Correct operands for a feature call)" I already found the reason for this issue: It was introduced with this change from Dec, 4th 2017 in class EQA_EXTRACTED_TEST_SET, feature "run_extracted_test" was changed in this way: https://svn.eiffel.com/viewvc/eiffelstudio/branches/Eiffel_18.07/Src/library/testing/eqa_extracted_test_set.e?r1=100315&r2=101093 Now, the implementation doesn't take into account the special notion for the object id in "a_operands" in the form of '#' + index (see comment on feature "context"). The previous implementation did this by calling "set_tuple_attributes (l_new, a_operands, 0)". I'm not sure about the correct implementation because simply reverting the change would ignore the previous FIXME comment in feature : -- FIXME: get rid of `empty_operands` usage. But for me reverting the change resulted in the expected result of the test execution. Maybe other changes introduced similar issues? I don't know... For more detail see here: https://groups.google.com/forum/#!topic/eiffel-users/TXnIeqHB0ck Best regards, Thomas
Create and execute an extracted test using AutoTest as described here: https://www.eiffel.org/doc/eiffelstudio/Using_extracted_tests The result is not as described in the tutorial but: "test_withdraw (TEST_BANK_ACCOUNT_002): FAIL (Correct operands for a feature call)"
Problem Report Interactions