Trinity is a small Tornado web application that provides a limited interface to a neo4j graph. The purpose is to offload the responsibilities of loading the graph from a main application process.
Trinity doesn't aim to be a gereral HTTP interface to neo4j - check out the neo4j REST server for that. This will offer a very specific, very optimized set of operations.
-
Python development headers
$ sudo apt-get install python-dev
-
Java Runtime Environment
$ sudo apt-get install default-jdk
-
jcc
$ sudo apt-get install jcc
OS X users have it easy, for once:
$ pip install jcc
- tornado
- JPype - http://sourceforge.net/projects/jpype/
- neo4j.py - http://components.neo4j.org/neo4j.py/
Install all of these using the pip-requirements.txt file with:
$ pip install -r pip-requirements.txt
The API endpoints are RESTful for the most part.
POST http://localhost:8888/node
{'id': 'bueda', 'node': {'username': 'bueda', 'user_id': 12345}}
POST http://localhost:8888/node/:id/relationships
{'to': ':other_id', 'data': {'other': 'data'}, 'type': 'MENTIONS'}
GET http://localhost:8888/node/:username/stats?stat=topics
› ./trinity.py --help
Usage: ./trinity.py [OPTIONS]
Options:
--help show this help information
--log_file_max_size max size of log files before rollover
--log_file_num_backups number of log files to keep
--log_file_prefix=PATH Path prefix for log files. Note that if you are running multiple tornado processes, log_file_prefix must be different for each of them (e.g. include the port number)
--log_to_stderr Send log output to stderr (colorized if possible). By default use stderr if --log_file_prefix is not set and no other logging is configured.
--logging=info|warning|error|none Set the Python log level. If 'none', tornado won't touch the logging configuration.
./trinity.py
--graph_path path to neo4j graph files
--port run on the given port
The default graph path is /var/neo4j, so make sure that path exists and is writable by your user.
To start the server:
› ./trinity.py