Open mecampbellsoup opened 4 years ago
BTW, I tried doing the following as well:
# Added to bottom of `api/v1/hosted_file_serializer.rb` file
API::V1::OutputSerializer = API::V1::HostedFileSerializer
I agree, It'd be great to specify the serializer in the method has_many and has_one.
like this:
has_one :foo, serializer API::V1::Foo
has_many :bar, serializer: API::V1::Bar
FWIW, I ended up defining Job#outputs
to return an array of HostedFile
instances, and this library seems to look at the type/class of the relationship... long story short, my JobSerializer#outputs
relationship is correctly serialized using my HostedFileSerializer
.
I have a
JobSerializer
that has 2 relationships - the hosted file, and outputs.(This is because a job is rendered from a file that the user uploaded - the "hosted file" - and the result of running the job are output files - "outputs".)
Both the singular hosted file, and the plural output, relationships are modeled by a class called
HostedFile
(not backed byActiveRecord
, but rather backed by the filesystem).I want to use the same serializer for both relationships, i.e.
API::V1::HostedFileSerializer
.Is it possible to specify the serializer to use for a relationship, e.g.:
Ideally I could then do something like the following and both relationships would be serialized using the
API::V1::HostedFileSerializer
: