GraphRepo is a tool for mining software repositories in real time. It indexes Git repositories in Neo4j and implements multiple queries to select and process the repository data.

For a complete description, see the online documentation.

The only requirement is to have Python >=3.5 and Docker installed on your system.

The production release can be installed using pip:

$ pip install graphrepo

The following instructions assume the Docker daemon is running on your machine:

$ docker run -p 7474:7474 -p 7687:7687 -v $HOME/neo4j/data:/data -v $HOME/neo4j/plugins:/plugins  -e NEO4JLABS_PLUGINS=\[\"apoc\"\]   -e NEO4J_AUTH=neo4j/neo4jj neo4j:3.5.11

Open a browser window and go to http://localhost:7474. Here you can configure the neo4j password. The default one is neo4jj.

--env NEO4J_dbms_memory_pagecache_size=4g
--env NEO4J_dbms_memory_heap_max__size=4g

In order to index a repository, you must clone it on localhost, and point GraphRepo to it. For example:

$ mkdir repos
$ cd repos
$ git clone https://github.com/ishepard/pydriller

Now enter the examples folder from this repository, and edit the configuration file for PyDriller to reflect the database URL and desired batch size:

$ cd ../examples/
$ nano configs/pydriller.yml

Afterwards, we can run the script from the examples folder which indexes the repository in Neo4j:

$ python -m examples.index_all --config=examples/configs/pydriller.yml

Go to http://localhost:7474 and use the query from 3.1

Assuming you succeded in step 1.4, use the follwing command to retrieve all indexed data:

$ python -m examples.mine_all --config=examples/configs/pydriller.yml

For a comprehensive introduction and more examples, see the documentation.

MATCH (n) RETURN n

MATCH (n) DETACH DELETE n;

MATCH (n:Commit)
// Take the first 100 commits nodes and their rels
WITH n LIMIT 100
DETACH DELETE n
RETURN count(*);

This project is enabled by Pydriller.