Open smaurel opened 5 years ago
The bulk result structure does not contain UpsertedIDs, so it is impossible to know what IDs were inserted when upserting using bulk upsert.
The fact that this feature is missing is a problem for requests such as the following one, when I want to do something with IDs that were inserted.
db.my_collection.bulkWrite([{"updateOne": {"filter": {"not_id_field" : "10000000"}, "update": {"$setOnInsert": {"_id": "some_random_id"},"$addToSet": {"other_field": {"$each": ["100"]}}}, "upsert": true }}])
{ "acknowledged" : true, "deletedCount" : 0, "insertedCount" : 0, "matchedCount" : 0, "upsertedCount" : 1, "insertedIds" : {}, "upsertedIds" : { "0" : "some_random_id" } }
Request result comes from Mongo 4.0.6 EDIT: The request result is the same using Mongo 3.6.8
Wouldn't it be more complicated than just adding the upsertedIDs field in BulkUpdate structure and using the field in the Bulk's runUpdate method?
The bulk result structure does not contain UpsertedIDs, so it is impossible to know what IDs were inserted when upserting using bulk upsert.
The fact that this feature is missing is a problem for requests such as the following one, when I want to do something with IDs that were inserted.
db.my_collection.bulkWrite([{"updateOne": {"filter": {"not_id_field" : "10000000"}, "update": {"$setOnInsert": {"_id": "some_random_id"},"$addToSet": {"other_field": {"$each": ["100"]}}}, "upsert": true }}]){ "acknowledged" : true, "deletedCount" : 0, "insertedCount" : 0, "matchedCount" : 0, "upsertedCount" : 1, "insertedIds" : {}, "upsertedIds" : { "0" : "some_random_id" } }Request result comes from Mongo 4.0.6 EDIT: The request result is the same using Mongo 3.6.8
Wouldn't it be more complicated than just adding the upsertedIDs field in BulkUpdate structure and using the field in the Bulk's runUpdate method?