SelfAPR: Self-supervised Program Repair with Test Execution Diagnostics

@inproceedings{selfAPR2022,
 title = {SelfAPR: Self-supervised Program Repair with Test Execution Diagnostics},
 author = {Ye, He and Martinez, Matias and Luo, Xiapu and Zhang, Tao and Monperrus, Martin},
 year = {2022},
 booktitle = {37th IEEE/ACM International Conference on Automated Software Engineering},
 articleno = {92},
 numpages = {13},
 publisher = {Association for Computing Machinery},
 doi = {10.1145/3551349.3556926}
}

Folder Structure

├── perturbation_model: the source code of java implemented perturbation model
│ 
├── Samples_BugLab: the generated perturbation-based samples with four BugLab repair actions
│ 
├── Samples_SelfAPR: the generated perturbation-based samples with 16 SelfAPR repair actions
│   
├── Dataset
│   ├── SelfAPR.csv.tar.gz : the dataset used to train SelfAPR model
│   ├── SelfAPR_ALL.csv.tar.gz : all generated training samples
│   ├── SelfAPR_FE.csv.tar.gz  : functional error training samples
│   ├── SelfAPR_CE.csv.tar.gz : compilation error training samples
│   ├── test.csv : the testing set from Defects4j bug dataset
│ 
├── result 
│   ├──defects4j-patches.txt : the correctly generated patches by SelfAPR trained model.
│   ├──patch_execution_result.csv : the execution results for patches generated for test set.
│   ├──valid_patches.csv : the patches classified as plausible or identical to the human-written patches.
└──  

Prerequisites:

Install Defects4J from https://github.com/rjust/defects4j 
export PATH=$PATH:"path2defects4j"/framework/bin

JDK 1.8 for Defects4J
export JAVA_HOME=/usr/lib/jvm/java-1.8.0-openjdk-amd64
export PATH=$JAVA_HOME/bin:$PATH
export CLASSPATH=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar

For python, install the prerequisities with:

pip3 install -r requirements.txt

Building the Java code

Check the source code and build the package as below.

cd perturbation_model
mvn package assembly:single

Please check perturbation-0.0.1-SNAPSHOT-jar-with-dependencies.jar under perturbation_model/target folder. Simplely check it with the different options:

java -jar perturbation-0.0.1-SNAPSHOT-jar-with-dependencies.jar path Options=SelfAPR|BugLab|test

With option SelfAPR: we perturb with all 16 rules in SelfAPR (Then the perturbation-based samples are executed).
With option BugLab: we perturb with all 16 rules in BugLab (No execution will be conducted).
With option test: we extract context information for testing samples.

We also uploaded a jar package on Zenodo. Download the jar file and execute it as below.

https://doi.org/10.5281/zenodo.6582348
java -jar ./perturbation_model/target/perturbation-0.0.1-SNAPSHOT-jar-with-dependencies.jar Your/JAVA/FILE/PATH Options=SelfAPR|BugLab|test-'buggyLineNo'

Code perturbation scripts

Checkout the buggy projects, apply the human written patches on them and make sure NO failling tests. Start the perturbation with this script:

python3  1_perturb_projects.py

Iterate each file of the considered projects, generate perturbed project-specific trianings samples and execute them against test cases:

python3  2_execute_perturbation.py

Prepare a set of evaluation bugs from Defects4J:

python3 3_prepare_test_data.py

We are ready to train the perturbed samples with transformer:Pytorch==1.7.1 and transformers>=4.10.0

pip install transformers
pip install sentencepiece
python3 4_train.py

To test the trained model:

python3 5_test.py

To evaluat the test results:

python3 6_evaluate_patch.py

Fault Localization with Gzoltar and Flacoco: faultlocalization

We compute all the suspicous buggy lines with two fault localization with Gzoltar and Flacoco. At inference phase, for a given suspicious statement found by fault localization tools, SelfAPR represents it with a sequences of tokens. Those tokens are given to the trained SelfAPR model. SelfAPR is configured by the inference beam size n, it outputs the n best patches for that suspicious statement.

All perturbed rules generated by SelfAPR

Perturb Actions	Explanation
P1 Replace type	modify declaring type ...
P2 Replace operator	modify operator ==, !=, etc
P3 replace literal	modify literal, "STRING", true, false
P4 replace constructor	modify constructor
P5 replace variable	modify arguments/swap argumens
P6 replace condition expression	reduce/expand boolean expression
P7 replace invocation	modify invocation
P8 compound statement	compound of rules
P9 replace with similarity	replace by transplanting a similar donor statement
P10 move statement	move a later statement before the target statement
P11 transplatant statement	transplanting a donor statement
P12 transplatant condition	wrap target statement with an existing conditional block
P13 transplatant a block	insert an existing block (if, loop, etc)
P14 remove a statement	delete statement
P15 unwrap a statement	unwrap block
P16 remove a block	remove block