fix(java): Fix usage of cursor pagination causing endpoint return type to be void #5708
Conversation
…n the request body
.../java/sdk/src/main/java/com/fern/java/client/generators/endpoint/AbstractEndpointWriter.java
Show resolved
Hide resolved
...ors/java/sdk/src/main/java/com/fern/java/client/generators/endpoint/PaginationPathUtils.java
Show resolved
Hide resolved
| return reference.visit(new UnboxedTypeNameGetter(generatorContext, reference)); | ||
| } | ||
|
|
||
| public static List<EnrichedCursorPathSetter> getPathSetters( |
There was a problem hiding this comment.
This function is quite long, so it could probably benefit from a variety of helper functions scoped to individual control flow statements.
It's fine to leave this as-is for now, but potentially something for us to revisit when we include those IR improvements to simplify this whole process.
There was a problem hiding this comment.
I'll make a linear ticket and put it on backlog
...ors/java/sdk/src/main/java/com/fern/java/client/generators/endpoint/PaginationPathUtils.java
Outdated
Show resolved
Hide resolved
| @@ -125,3 +127,4 @@ allowedFailures: | |||
| - unions | |||
| - streaming-parameter | |||
| - literal | |||
| - java-pagination-deep-cursor-path | |||
There was a problem hiding this comment.
Do you think we still need this test definition alongside the testdata already included in the pagination testdata?
I think we have a lot of nested cursor pagination examples there, but feel free to leave this as-is if I'm wrong. If not here, wventually we'll want to consolidate the endpoints here in the pagination test definition so everyone can benefit.
There was a problem hiding this comment.
I'd rather leave it as is until the java one deals with these cases
This PR addresses a use case where a complex return type will make cursor paginated endpoints return void. PR description update pending more investigation.