HTTP endpoint sink¶
This is a special data sink that registers an HTTP publisher endpoint that one can get entities from.
A pipe that references the HTTP endpoint
sink will not pump any
entities, in practice this means that a pump is not configured for the
pipe; the only way for entities to flow through the pipe is by
retrieving them from the HTTP endpoint.
It exposes these URLs:
URL |
Description |
---|---|
|
JSON entities endpoint |
|
JSON entities endpoint - filename in URL variant |
|
SDShare collections feed |
|
SDShare fragments feed |
The serialisation of entities as JSON is described in more detail here. This endpoint is compatible with The JSON source.
Note that any URL parameters given to these endpoints are bound to a DTL variable named _B
and is available to any DTL transform on the pipe in which the endpoint sink is a part, see
DTL Variables for more details.
The SDShare protocol is described here.
The exposed URLs may support additional parameters such as since
and limit
- see
the API reference for the full details.
Prototype¶
{
"type": "http_endpoint"
}
Properties¶
Property |
Type |
Description |
Default |
Req |
---|---|---|---|---|
|
String |
This property provides a hint to HTTP clients on what filename to use when downloading data (via the
|
||
|
String |
This property provides a hint to HTTP clients how to render the file data. The valid values are |
|
Example configuration¶
The pipe configuration given below will expose the my-entities
publisher endpoint and read the entities from the my-entities
dataset
{
"_id": "my-entities",
"name": "My published entities endpoint",
"type": "pipe",
"sink": {
"type": "http_endpoint"
}
}