Streams

Streams are complimentary to actions. Think of streams as views in SQL. A stream is basically one entity + set of transformations and filters on the entity. Streams are read-only and exposed with similar semantics of that of entities. Daptin will expose JSONAPI for each stream just like it does for entities.

Here is an example of a stream which exposes list of completed todos only

{
        StreamName:     "transformed_user",
        RootEntityName: "todo",
        Columns: []api2go.ColumnInfo{          // List of columns in this stream
            {
                Name:       "transformed_todo_title",  
                ColumnType: "label",
            },
            {
                Name:       "completed_on",
                ColumnType: "datetime",
            },
        },
        QueryParams: QueryParams{
            "Filter": "completed=true",
            "Select": "title,deadline",
        },
        Transformations: []Transformation{
            {
                Operation: "select",
                Attributes: map[string]interface{}{
                    "columns": []string{"title", "deadline"},
                },
            },
            {
                Operation: "rename",
                Attributes: map[string]interface{}{
                    "oldName": "title",
                    "newName": "transformed_todo_title",
                },
            },
            {
                Operation: "rename",
                Attributes: map[string]interface{}{
                    "oldName": "deadline",
                    "newName": "completed_on",
                },
            },
        },
}   

Daptin uses the library kniren/gota to systematically specific list of transformations which are applied to the original data stream.