Add nullability support for server parameter delegation

[nomisRev]

Subsystem
Server, core

Motivation
Currently it is not possible to use nullable values with parameter delegation.

get("/nullable") {
    val nameOrNull: String? by call.queryParameters
    val ageOrNull: Int? by parameters
}

Solution

This PR add support by removing the : Any constraint and relying on typeInfo.kotlinType?.isMarkedNullable to check if null is allowed or not.

[nomisRev]

I'm a bit unsure how reliable typeInfo.kotlinType?.isMarkedNullable is across platforms?

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: e0610b9a-2abc-4ef8-a8f8-deab6e4b869b

📥 Commits

Reviewing files that changed from the base of the PR and between 86d7c3f and 1056631.

📒 Files selected for processing (1)

• ktor-server/ktor-server-core/common/src/io/ktor/server/util/Parameters.kt

🚧 Files skipped from review as they are similar to previous changes (1)

• ktor-server/ktor-server-core/common/src/io/ktor/server/util/Parameters.kt

---

📝 Walkthrough

Walkthrough

Generic type constraints for Parameters delegated accessors were relaxed to allow nullable result types; implementation now returns null for absent parameters when the target type is nullable and preserves existing exceptions for non-nullable types and conversion failures.

Changes

Cohort / File(s)	Summary
API Signatures ktor-server/ktor-server-core/api/ktor-server-core.klib.api	Adjusted exported signatures: getOrFailImpl and getValue type parameter bounds changed from A: Any to A: Any? (nullable-capable generics).
Core Implementation ktor-server/ktor-server-core/common/src/io/ktor/server/util/Parameters.kt	Removed non-null constraint on R for getValue and getOrFail; getOrFailImpl now checks typeInfo.kotlinType?.isMarkedNullable and returns null as R when appropriate, otherwise throws MissingRequestParameterException or ParameterConversionException as before. KDoc and expression-bodied forms updated.
Tests ktor-server/ktor-server-core/common/test/io/ktor/server/http/ParametersTest.kt	Added tests covering nullable delegated properties resolving to value or null, conversion failure for nullable target, and missing non-null delegated value throwing MissingRequestParameterException.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 11.11% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and concisely summarizes the main change: adding nullability support for server parameter delegation, which is the core objective of the PR.
Description check	✅ Passed	The description covers all required template sections with relevant detail: subsystem (Server, core), motivation (inability to use nullable values with delegation), and solution (removing : Any constraint and checking nullability).

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch nomisrev/nullable-parameter-delegation

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[bjhham]

Looks like an easy win 👍

I'm pretty sure isMarkedNullable works cross-platform on reified KTypes.

I don't know why we didn't just implement it that way in the first place...

[bjhham]

FYI logged an issue to handle property delegation in the compiler plugin:
KTOR-9422 OpenAPI code inference misses property delegation

[e5l]

lgtm

[nomisRev]

@osipxd I updated the PR to move the nullability check in the actual implementation. This way we now also support nullable types for getOrFail.

TypeInfo is always being created because it's used by the underlying conversions mechanism. Perhaps the conversions mechanism can be optimised to avoid TypeInfo where possible but I left that for outside of this scope since it seems unrelated to my changes.

Could you review again? Thank you for the careful review and pointing it out! 🙏 We're still getting a better outcome thanks to it!

[osipxd]

Thank you for updating the PR!
Only one thing I've noticed. Compilation fails here: ‎property is unresolved in this scope (probably should be ‎name) 👀

[nomisRev]

Yes, sorry about that! I was coding without loading the project 😅 Takes too much resources 💀