Optimize routing resolution allocations with state machine refactoring

[e5l]

Reduce per-request allocations in RoutingResolveContext by converting from parameter/return recursion to field-based state machine with Unit return type. Replace ArrayList with parallel arrays for backtracking state. Optimize tailcard selector to eliminate intermediate list allocations.

• Convert handleRoute to Unit return, eliminate Double boxing from suspend
• Use parallel arrays (traitRoutes, traitParameters, traitQualities) for state
• Reuse ParametersBuilder field to avoid allocation in findBestRoute
• Create Success objects only when tracing enabled
• Optimize PathSegmentTailcardRouteSelector to single-pass iteration
• All algorithms and semantics preserved; all tests passing

[coderabbitai]

Important

Review skipped

Draft detected.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

---

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.}

[Copilot]

Pull request overview

This PR optimizes routing resolution in Ktor by reducing per-request memory allocations through a state machine refactoring. The optimization converts recursive parameter passing to field-based state management and replaces object allocations with primitive array operations.

Changes:

• Converted handleRoute from returning Double quality values to using a field-based state machine with Unit return type
• Replaced ArrayList<Success> with parallel arrays (traitRoutes, traitParameters, traitQualities) to eliminate Success object allocations during backtracking
• Optimized PathSegmentTailcardRouteSelector to use single-pass iteration instead of drop().mapIndexed() chains

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
ktor-server/ktor-server-core/common/src/io/ktor/server/routing/RoutingResolveContext.kt	Core refactoring from recursive return values to state machine using parallel arrays for backtracking state; Success objects now created only when tracing is enabled
ktor-server/ktor-server-core/common/src/io/ktor/server/routing/RouteSelector.kt	Optimized PathSegmentTailcardRouteSelector to avoid intermediate list allocations using direct ArrayList construction

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[bjhham]

Looks like there's an increase in RoutingResolveContext and we may have missed a HttpProtocolVersion increase from earlier changes due to leniency in the assertion.

    Request consumes 20.24 KB, expected 18.81 KB. 
      Difference: 1464 < 2311 (allowed)
      Consumed 20.24 KB on request
      Expected 18.81 KB on request
      Extra    1464 bytes on request
    (See stdout + build/allocations/* files for details)
    
    Increased locations:
      HttpProtocolVersion.kt                         816    (816          -->         1632)
      RoutingResolveContext.kt                       632    (1432         -->         2064)
      RoutingRoot.kt                                  40    (344          -->          384)
      BaseApplicationEngine.kt                        16    (248          -->          264)
      BaseApplicationResponse.kt                       8    (248          -->          256)
    
    Decreased locations:
      RouteSelector.kt                               -48    (128          -->           80)

[e5l]

thanks, let me investigate other sites as well

[bjhham]

Allocations are looking pretty good now 😄