aidarvaleev1998/method-embedding

Source Code Tools

Library for analyzing source code with graphs and NLP. What this repository can do:

1. Fetch source codes for packages in pip
2. Create indexes of python packages using Sourcetrail
3. Convert Sourcetrail indexes into a connected graph
4. Build graphs for source codes from AST
5. Train Graph Neural Network for learning representations for source code
6. Predict Python types using NLP and graph embeddings

For more details consult our wiki.

Installing Python Libraries

You need to use conda, create virtual environment SourceCodeTools with python 3.8

conda create -n SourceCodeTools python=3.8

(optional) If you plan to use graphviz

conda install -c conda-forge pygraphviz graphviz

(optional) Install CUDA 11.1 if needed

conda install -c nvidia cudatoolkit=11.1

To install SourceCodeTools library run

git clone https://github.com/VitalyRomanov/method-embedding.git
cd method-embedding
pip install -e .
# pip install -e .[gpu]

In case there are errors during installing dgl package install the dependency manually

pip install dgl==0.7.1 -f https://data.dgl.ai/wheels/repo.html
#pip install dgl-cu111==0.7.1 -f https://data.dgl.ai/wheels/repo.html

Processing Python Code

Source code should be structured in the following way

source_code_data    
│
└───package1
│   │───source_file_1.py
│   │───source_file_2.py
│   └───subfolder_if_needed
│       │───source_file_3.py
│       └───source_file_4.py
│   
└───package2
    │───source_file_1.py
    └───source_file_2.py

An example of source code data can be found in this repository method-embedding\res\python_testdata\example_code. A package should contain self-sufficient code with its dependencies. Unmet dependencies will be labeled as non-indexed symbol.

Indexing with Docker

To create dataset need to first perform indexing with Sourcetrail. The easiest way to do this is with a docker container

docker run -it -v "/full/path/to/data/folder":/dataset mortiv16/sourcetrail_indexer

Creating graph

Need to provide a sentencepiece model for subtokenization. Model trained on CodeSearchNet can be downloaded here.

SCT=/path/to/SourceCodeTool_repository
SOURCE_CODE=/path/to/source/code/indexed/with/sourcetrail
DATASET_OUTPUT=/path/to/dataset/output
python $SCT/SourceCodeTools/code/data/sourcetrail/DatasetCreator2.py --bpe_tokenizer sentencepiece_bpe.model --track_offsets --do_extraction $SOURCE_CODE $DATASET_OUTPUT

The graph dataset format is described in wiki

graph_dataset    
│
└───no_ast
│   │───common_call_seq.bz2
│   │───common_edges.bz2
│   │───common_function_variable_pairs.bz2
│   │───common_nodes.bz2
│   │───common_source_graph_bodies.bz2
│   └───node_names.bz2
│   
└───with_ast
    │───common_call_seq.bz2
    │───common_edges.bz2
    │───common_function_variable_pairs.bz2
    │───common_nodes.bz2
    │───common_source_graph_bodies.bz2
    └───node_names.bz2

no_ast contains graph built from global relationships only. with_ast contains graph with AST nodes and edges. Two main files for building the graph are common_nodes.bz2 and common_edges.bz2. The files are stored as pickled pandas table (read with pandas.read_pickle) and probably not portable between platforms. One can view the content by converting table into the csv format

python $SCT/SourceCodeTools/code/data/sourcetrail/pandas_format_converter.py common_nodes.bz2 csv

The graph data can be loaded as pandas tables using load_data function

from SourceCodeTools.code.data.dataset.Dataset import load_data

nodes, edges = load_data(
    node_path="path/to/common_nodes.bz2",
    edge_path="path/to/common_edges.bz2"
)