gremlin-python-example

An example project for gremlin_python, as there is very little public code out there using it.

This project also uses Chalice (aws/chalice)

Instructions

>> mkvirtualenv neptunedemo --python=python3.6
>> pip install chalice
>> pip install gremlinpython

• Clone this repo to your project folder
• Set GRAPH_DB in .chalice/config.json to your graph db
• export AWS_DEFAULT_REGION=us-east-1 if using AWS Neptune

>> chalice deploy

Then send http requests (examples below) to your new endpoint.

If using AWS Neptune:

• Add AWSLambdaVPCAccessExecutionRole to the lambda role
• Set the correct VPC, subnet and SG

Example Requests

New Person

POST to https://whatever.execute-api.us-east-1.amazonaws.com/api/person

{"id": "XYZ1234", "prop1": "Some", "prop2": "Value"}

Retrieve Person

GET to https://whatever.execute-api.us-east-1.amazonaws.com/api/person/XYZ1234

Update Person

PUT to https://whatever.execute-api.us-east-1.amazonaws.com/api/person/XYZ1234

{"id": "XYZ1234", "prop1": "Some", "prop2": "NewValue"}

Upsert Relationship

POST to https://whatever.execute-api.us-east-1.amazonaws.com/api/relationship

{"from": "XYZ1234", "to": "BOB", "weight": "1.0"}

Retrieve Known Associates

GET to https://whatever.execute-api.us-east-1.amazonaws.com/api/relationship/XYZ1234

You can optionally add a ?threshold=0.n parameter to the querystring to set the threshold of relationship weighting to retrieve.

Clear Graph

DELETE to https://whatever.execute-api.us-east-1.amazonaws.com/api/clear

TODOs

• Cleanup the retrieval query
• Add more unit tests using TinkerGraph
• Add Cognito for authorisation of requests
• Extend demo instructions to include graph db setup
• Add additional options to the search

Additional reading

• http://kelvinlawrence.net/book/Gremlin-Graph-Guide.html
• https://docs.aws.amazon.com/neptune/latest/userguide/
• http://chalice.readthedocs.io/