Important (April 13, 2012): Small changes have been made to the packaging of the code submitted along with the FSE2012 paper. Please see artifacts.txt for more information.
Almost as important: The rest of this document is out of date.
This project uses Apache Maven to manage all dependencies and versioning. The simplest way to get going is to run the following command: mvn clean compile package assembly:single
This will clean the source tree, compile the code, run the tests, package the code into a jar file, and finally copy all the libraries to a location that makes some modicum of sense. Then to run the mining scripts just run: ./github.sh
Project configuration is controlled through the configuration.properties
file in src/main/resources. Don't commit it with your github username
and apitoken, that would be bad.
Actually, right now it doesn't actually use those fields and it probably won't anytime in the future. So don't worry so much about that.
Various elements of the miner can be turned on and off by changing the values of their field to anything other than true. Those fields are:
- net.wagstrom.research.github.miner.issues
- net.wagstrom.research.github.miner.gists
- net.wagstrom.research.github.miner.repositories
- net.wagstrom.research.github.miner.organizations
- net.wagstrom.research.github.miner.users
Every Vertex in the database should have the following:
- type: one of USER, REPOSITORY
- created_at: ISO 8601 formatted date of when the node was created
Every Edge in the database should have the followning fields:
- label: not really a field, but always present
- created_at: ISO 8601 formatted date of when the edge was created