https://github.com/mohannadcse/normalizer-authorship

https://github.com/mohannadcse/normalizer-authorship

Last synced: 4 days ago
JSON representation

• Host: GitHub
• URL: https://github.com/mohannadcse/normalizer-authorship
• Owner: Mohannadcse
• License: mit
• Created: 2022-10-02T17:14:16.000Z (about 2 years ago)
• Default Branch: main
• Last Pushed: 2023-07-18T17:02:40.000Z (over 1 year ago)
• Last Synced: 2024-12-06T15:50:12.576Z (20 days ago)
• Language: C++
• Size: 407 KB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
## The normalizer implementation consists of 2 components: 

1- [normalized transformations](src/LibToolingAST_Normalizer)

2- [generating normalized Dataset](src/PyProject/normalizer)

## Preparing Projects

### Clang Transformations

Follow the instructions [here](https://github.com/EQuiw/code-imitator/blob/master/src/LibToolingAST/README.md)

### Python Project

 

- Export the python project path `i.e., export PYTHONPATH="/nobackup/code-imitator-normalizer/src/PyProject:$PYTHONPATH"`

- Follow the instructions [here](https://github.com/EQuiw/code-imitator/blob/master/src/PyProject/README.md)

## Dataset

The dataset shold follow the same structure of the original dataset https://github.com/EQuiw/code-imitator/tree/master/data/dataset_2017/dataset_2017_8

## Running the Normalizer

 1- Download the [dataset](https://github.com/EQuiw/code-imitator/tree/master/data/dataset_2017/dataset_2017_8)

 2- Specifying the location of the `datasetpath` and `normalizedresultspath` inside the python file `src/PyProject/Configuration.py`

 3- Specify the attack configuration as defined [here](src/PyProject/Configuration.py#L90)

 4- Provide the corresponding pickle files (classification models) to the specified attack in the previous step. The default location is defined [here](src/PyProject/Configuration.py#L30)

 

 5- The verification of the output before and after the applying the transformer can be enable/disabled from [here](src/PyProject/Configuration.py#L96). It's by default enabled. 

 6- To run the normalizer and generate the normalized dataset `python3 src/PyProject/normalizer/normalizer.py `, where the current implementation supports the following categories: `API, Control, Misc, Declaration, all`

## Ouput of the Normalizer

Besides generating the normalized files for the provided dataset in the specified location `normalizedresultspath`, the normalizer also generates two CSV files `tranfromation_results.csv` and `count_results.csv` to report about the performed transformations for each `cpp` file.

# Debugging the Normalizer

We added debugging feature to the normalizer to normalize just a signle `.cpp` file. The variable `single` inside the python file `src/PyProject/Configuration.py` enables this debugging. Otherwise, it should be zero.

# Disclaimer

The normalizer has been developed to normalize code transformations defined [here](https://github.com/EQuiw/code-imitator/tree/master/src/LibToolingAST). Therefore, many of its libraries have been used as is or adapted.

# Citation 

If you are using our implementation, please cite our NeurIPS '22 paper. You may use the following BibTex entry:

```

@inproceedings{

wang2022robust,

title={Robust Learning against Relational Adversaries},

author={Yizhen Wang and Mohannad Alhanahnah and Xiaozhu Meng and Ke Wang and Mihai Christodorescu and Somesh Jha},

booktitle={Advances in Neural Information Processing Systems},

editor={Alice H. Oh and Alekh Agarwal and Danielle Belgrave and Kyunghyun Cho},

year={2022},

url={https://openreview.net/forum?id=WBp4dli3No6}

}

```