You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have a usecase where I want to use the multisearch API to search two separate indices, with differing document types. Looks like this was quite straightforward using NEST, but I can't find an easy way using Elastic.Clients.Elasticsearch 8.12.0. The type argument that is present on all the MultiSearch methods implies that returning different document types for each search is not supported. Is that correct?
I suppose I could specify TDocument as object or string and do my own serialization, but would much prefer out-of-the-box support for this. Is that something you plan to provide in the future?
The text was updated successfully, but these errors were encountered:
this is currently not on my roadmap, but sounds like a useful feature. I have to check how this was implemented in NEST and how much effort it would be to port it over. Feel free to create a PR as well!
Deserializing to object or RawJsonString followed by custom deserialization call should work for now.
An alternative would be to use polymorphism with System.Text.Json. This option however, would probably require you to add a new property to your already ingested data.
I have a usecase where I want to use the multisearch API to search two separate indices, with differing document types. Looks like this was quite straightforward using NEST, but I can't find an easy way using
Elastic.Clients.Elasticsearch 8.12.0
. The type argument that is present on all theMultiSearch
methods implies that returning different document types for each search is not supported. Is that correct?I suppose I could specify
TDocument
asobject
orstring
and do my own serialization, but would much prefer out-of-the-box support for this. Is that something you plan to provide in the future?The text was updated successfully, but these errors were encountered: