Remove relative path in Dspot InputConfiguration

[nicolabertazzo]

Characteristics

• Issue Type: [feature]
• Severity: [feature]
• Reporter: [Nicola Bertazzo, [email protected]]

Description

In dspot/dspot/src/main/java/fr/inria/diversify/utils/sosiefier/InputConfiguration.java the file path of some properties (testSrc, src, classes, tmpDir) are stored as relative path (String). My proposal is to store these properties as absolute path or File.

Help on issue template

Preview to follow the link or open file .github/ISSUE_DOC.md

[danglotb]

Hi @nicolabertazzo

Are you working on this?

I could have a look, I'll refactor InputConfiguration and InputProgram objects to unify them.

[nicolabertazzo]

No i'm fixing some problem about maven plugin (adding the new in line parameters).
I'll update InputConfiguration before the end of this week.

[nicolabertazzo]

Hi @danglotb ,
I fixed some problems about inputConfiguration refactoring and I pushed to master.
I have an issue when i run dspot from maven on dhell.
If i execute from dhell folder the follow command:

mvn clean test eu.stamp-project:dspot-maven:1.1.1-SNAPSHOT:amplify-unit-tests -Diteration=3 -Dtests=eu.stampproject.examples.dhell.HelloAppTest -Dfilter=eu.stamp_project.examples.dhell*

The result is:

The original test suite kill 114 / 371

But if i execute dspot from pit using this command:

mvn org.pitest:pitest-maven:1.3.0:mutationCoverage -DwithHistory -DtargetClasses=eu.stamp_project.examples.dhell* -DreportsDirectory=target/pit-reports -DoutputFormats=CSV,HTML -DtimeoutConst=10000 -DjvmArgs=16G -DmutationEngines=gregor -Dmutators=ALL -DadditionalClasspathElements=target/dspot/dependencies/

the result is different.

@danglotb can you check if dspot runs correctly from maven?

[danglotb]

the result is different.

What do you mean? The original test suite does not kill 114 mutants over 371?

I fixed some problems about inputConfiguration refactoring and I pushed to master.

please, do not push directly on master, but rather open a pull request.

[nicolabertazzo]

The original test suite generate this result:
`>> Generated 371 mutations Killed 120 (32%)

Ran 5 tests (0.01 tests per mutation)
=========================================`

I'm sorry if I pushed into the master. It was a mistake.

[danglotb]

The difference is not significative. Thus, the number of generates mutants is the same.

The difference in the killed mutant can come from TIMED_OUT execution or MEMORY errors.

It would be great if you can provide the .csv files.

[danglotb]

This has been fixed with recent changes.