Executer

In general a workflow is a graph consists of operators as nodes and links as connection between the operators. The definition of the workflow datastructure is used to transfer workflows between the components. We would like to use a simplified version in the first development stage and later use a more complex version. It consists of a triple of the identifier of data source, extractor and target graph.

Simple example

This example of a workflow contains three operators (data, extractor, targetgraph) and two links (data -> extractor -> targetgraph) between them. This format is restricted to three operators. The data node is start of the workflow and the targetgraph is the end.

{
    "data" : "file",
    "extractor" : "fox",
    "targetgraph" : "thegraph"
}

Complex example

This example of a workflow contains two operators (id_1, id_2) and an links (id_1 -> id_2) between them. In this way we are not restricted, e.g. we can create forks or multiple starts.

{
    "links":[
        {
            "fromConnector":"A",
            "toConnector":"B",
            "fromOperator":"id_1",
            "toOperator":"id_2"
        }
    ],
    "operators":{
        "id_1":{
            "properties":{
                "id":"id_1",
                "content":"/text.txt",
                "type":"file",
                "outputs":{
                    "A":{
                        "label":"NL"
                    }
                }
            }
        },
        "id_2":{
            "properties":{
                "id":"id_2",
                "content":"FRED-MS",
                "type":"extractor",
                "input":{
                    "B":{
                        "label":"RDF"
                    }
                }
            }
        }
    }
}