CODEWITHSUNDEEP
pythongoogle-app-engineasynchronousapp-engine-ndb
Jon
Jon

Reputation: 794

ndb.Tasklet and SyncTasklet with Multiple Nested Functions

I am trying to efficiently call tasklet and subtasklets:

@ndb.tasklet
def getBeds(bed_key):
    bed = yield bed_key.get_asyn()
    bed_info = {}
    ......
    raise ndb.Return(bed_info)

@ndb.tasklet
def getRoom(room_key):
    room = yield room_key.get_async()
    room_info = {}
    ..........
    beds_in_room = map(getBeds,room.beds)
    room_info["beds"] = beds_in_room
    raise ndb.Return(room_info)

@ndb.tasklet
def getBuilding(build_key):
    build = yield build_key.get_async()
    build_info = {}
    ...........
    rooms_in_build = map(getRoom,build.rooms)
    build_info["rooms"] = rooms_in_build

@ndb.toplevel
def getHotel(hotel_obj)
    hotel_inf = {}
    .........
    buildings_in_hotel = map(getBuilding,hotel_obj.buildings)
    hotel_inf["buildings"] = buildings_in_hotel
    return hotel_inf

For some reason, I though the @ndb.toplevel would pause getHotel until everything finished. Unfortunately, buildings_in_hotel is returning a list of futures...

How do I get it to complete?

Upvotes: 3

Views: 443

Answers (1)

tesdal
tesdal

Reputation: 2459

You yield!

When calling a tasklet you always get a future, then yield to get the result. There is also parallel yield, where you yield a tuple or list of tasklets (sometimes referred to as barrier).

beds_in_room = yield map(getBeds,room.beds)

and

rooms_in_build = yield map(getRoom,build.rooms)

and

buildings_in_hotel = yield map(getBuilding,hotel_obj.buildings)

Upvotes: 3

Related Questions