Are Nopol and Dynamoth patches all correct?

ASSERT-KTH / drr

Tool & data on the correctness of Defects4 patches generated by program repair tools http://arxiv.org/pdf/1909.13694

http://arxiv.org/pdf/1909.13694

Creative Commons Attribution Share Alike 4.0 International

10 stars 6 forks source link

Are Nopol and Dynamoth patches all correct? #24

Closed SophieHYe closed 6 years ago

SophieHYe commented 6 years ago

Hi @monperrus @martinezmatias @tdurieux

There are 103 nopol patches from result of March 2017: https://github.com/Spirals-Team/defects4j-repair/tree/master/results/2017-march and 153 Dynamoth patches from https://github.com/tdurieux/dynamoth-experiments/tree/master/results/2017-november.

I would like to know have they been manual analyzed as correct patches or plausible patches?

In addition, there is no patches from jkali, should we add them?

tdurieux commented 6 years ago

There is no human analysis for those patches

monperrus commented 6 years ago

It's an open question whether we should use them. We don't want to overflow the paper with patches of our own tools that are not formally part of peer-reviewed papers.