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
Currently, when the decoder encounters a schema whose type is an array (such as ["string", "null"]) is ignores the object and leave it entirely as is.
Continuing on my streak of implementing the bits of OpenAPI 3.1.0 that I encounter on the schema I'm working with, I'd like to handle cases where the Schema.type is an array of types. In my case, it's exclusively [something, "null"], but if can be anything. According to some sources, it looks like the expected behaviour is that the schema object can contain any restriction field that would make sense for any of the types in the array. The expected behaviour is similar to a corresponding anyOf.
I've only been looking at the decoding so far, and I got it working by merging all the various clauses that deal with various possibilities into one big function that convert any field it finds. This works well, except for one piece of logic which depends on the actual type: properties should be %{} instead of nil if the type is/contains object.
This solution isn't very satisfying, so I'll sleep on it, see if I can figure a nicer way. The current progress lives at #659.
The text was updated successfully, but these errors were encountered:
Currently, when the decoder encounters a schema whose type is an array (such as
["string", "null"]
) is ignores the object and leave it entirely as is.Continuing on my streak of implementing the bits of OpenAPI 3.1.0 that I encounter on the schema I'm working with, I'd like to handle cases where the
Schema.type
is an array of types. In my case, it's exclusively[something, "null"]
, but if can be anything. According to some sources, it looks like the expected behaviour is that the schema object can contain any restriction field that would make sense for any of the types in the array. The expected behaviour is similar to a correspondinganyOf
.I've only been looking at the decoding so far, and I got it working by merging all the various clauses that deal with various possibilities into one big function that convert any field it finds. This works well, except for one piece of logic which depends on the actual type:
properties
should be%{}
instead ofnil
if the type is/containsobject
.This solution isn't very satisfying, so I'll sleep on it, see if I can figure a nicer way. The current progress lives at #659.
The text was updated successfully, but these errors were encountered: