Code Semantic Enrichment for Deep Code Search

Tested in Ubuntu 16.04

• Python 2.7-3.6
• Keras 2.1.3 or newer
• Tensorflow-gpu 1.7.0
• lucene 7.7.1

The datasets used in our paper will be found at: https://drive.google.com/drive/folders/1j-0xukLQWGrJ8-Lxw7vFAbubFTyXJT2C?usp=sharing

If you want to reprocess the data, you can process it into a usable form for the model by following steps:

1.Build corpus for each features (i.e., description, tokens):

python createCorpus.py python createVocab.py python vocab2pkl.py

2.Processing training data and testing data according to the corpus:

python txt2pkl.py

Build retrieval base: python Index.py

Perform search: python Search.py

Remove stop words: python deleteStopWords.py

Put the data set into the data/github directory under keras

Edit hyper-parameters and settings in config.py

python main.py --mode train python main.py --mode eval