cheap-cheer / clbi-glance

Code Line-level Bugginess Identification

The replication kit of GLANCE

Title: Code-line-level bugginess identification: How far have we come, and how far have we yet to go?

This repository stores the source codes of the four categories of state-of-the-art CLBI approaches.

1. Folders Introduction

• CLBI/src/ This folder stores the source code of SATs, NLPs, MITs, and GLANCE written in Python.

• CLBI/result/ This folder stores all classification results of the each approaches.

• BugDet/dataset/ This folder stores all 19 projects in datasets.

2. Execution commands

In order to make it easier to obtain the classification results, one can run it according to the following command regulation.

python main.py [model_name]

In above command,

• [model_name] indicates a CLBI approach, i.e., PMD, CheckStyle, NGram, NGram_C, TMI_LR, TMI_SVM, TMI_MNB, TMI_DT, TMI_RF, LineDP, GLANCE_EA, GLANCE_MD, and GLANCE_LR.

Here is some usage examples:

python main.py PMD

python main.py GLANCE_EA

3. Studied Approaches

Approach	Source path	Description
SAT	src/models/tools.py	Static Analysis tools
NLP	src/models/natural.py	Natural language processing
MIT	src/models/explain.py	Model interpretation techniques
GLANCE	src/models/glance.py	Aiming at control- and complex-statements

4. Contact us

Mail: [email protected]