Graph-service

This is a service that provides a REST API (according to the HAL+JSON specification) against which you can create and query nodes and relationships in a graph.

The service uses neo4j version 3.5 internally to store data.

There is also a small client, that implements a (very poor) algorithm to calculate the shortest path between two nodes using the API. The algorithm is naively implemented and will not take edge cases and circular dependencies into consideration (with more time I would have adapted a proper algorithm such as Dijkstras).

To run the tests

First start a neo4j database using Docker:

docker run \
     --name testneo4j \
     -p7474:7474 -p7687:7687 \
     -d \
     -v $HOME/neo4j/data:/data \
     -v $HOME/neo4j/logs:/logs \
     -v $HOME/neo4j/import:/var/lib/neo4j/import \
     -v $HOME/neo4j/plugins:/plugins \
     --env NEO4J_AUTH=neo4j/test \
     neo4j:3.5

Then run tests by running lein test

The tests use the open source library halboy https://github.com/jimmythompson/halboy to traverse the API and its resources.

To run the service

First start a neo4j database using Docker:

docker run \
     --name testneo4j \
     -p7474:7474 -p7687:7687 \
     -d \
     -v $HOME/neo4j/data:/data \
     -v $HOME/neo4j/logs:/logs \
     -v $HOME/neo4j/import:/var/lib/neo4j/import \
     -v $HOME/neo4j/plugins:/plugins \
     --env NEO4J_AUTH=neo4j/test \
     neo4j:3.5

Then run lein run

The service will be available under http://localhost:1234

The API (mostly) conforms to the HAL+JSON specification so it's partially self-documenting and explorable.

Nodes require a JSON body but only optionally takes decorative fields under the key propeties, for example:

{
  "properties": {
    "someField": "someValue"
  }
}

Relationships require the fields from, to, type and optionally takes additional fields under the key propeties, for example:

{
  "from": "<SOME_NODE_ID>",
  "to": "<SOME_OTHER_NODE_ID>",
  "type": "friend",
  "properties": {
    "someField": "someValue"
  }
}

Due to time constraints there is no validation added and currently the running service and the component tests use the same underlying database, which of course you wouldn't do on an actual production service.