Async Version of annotate_note

[boyleconnor]

nlpsandbox-client/nlpsandboxclient/client.py

Line 344 in 540fbfe

def annotate_note(host: str, note: dict, annotator_type: str) -> dict:

It would be very useful to have an asynchronous version of annotate_note (as well as a few of the other methods in this module).

E.g. I call something like this:

requests = []
for annotation_type in ('person', 'date', 'address'):
    requests.append(annotate_note(note=my_note, annotator_type=annotation_type, host=HOST_URL))

responses = []
for request in requests:
    responses.append(request.response())

[thomasyu888]

@cascadianblue

When using the auto-generated client, there is a way to set async_req=True:

configuration = annotator.Configuration(host=host)
with annotator.ApiClient(configuration) as api_client:
    api_instance = text_date_annotation_api.TextDateAnnotationApi(api_client)
    annotations = api_instance.create_text_date_annotations(
        text_date_annotation_request=text_date_request,
        async_req=True
    )

The object returned annotations has 4 functions:

• annotations.get()
• annotations.ready()
• annotations.successful()
• annotations.wait()

If I create a parameter async, should I only return the annotations async object?

[tschaffter]

@thomasyu888 Is there an update regarding this ticket? Is this a low hanging fruit? Is it a must-have for the launch on June 8th?

[thomasyu888]

@tschaffter not a must have for launch, and there hasn't been much development on this front for awhile.