fix: sort ASCII-safe strings by runtime kind#868
Open
He-Pin wants to merge 1 commit into
Open
Conversation
Motivation: AsciiSafeStr is a Val.Str subclass used by renderer optimizations, but std.sort compared exact runtime classes. Sorting arrays containing ASCII-safe strings could fail even though Jsonnet semantics treat them as strings. Modification: Classify sortable values by Jsonnet runtime kind instead of exact JVM/Native class, preserving existing fast paths for strings, numbers, arrays, booleans, and objects. Add a regression covering all-AsciiSafeStr, mixed Str/AsciiSafeStr, and key-function-produced AsciiSafeStr sorting. Result: std.sort now treats AsciiSafeStr values as strings without changing Jsonnet semantics. Full cross-platform validation will be run before opening the draft PR.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Motivation:
Val.AsciiSafeStris aVal.Strsubclass used by renderer/string optimizations.std.sortwas using exact runtime class equality to validate and dispatch sortable values, so arrays containingAsciiSafeStrcould fail even though Jsonnet semantics treat those values as strings.Key Design Decision:
Classify values by Jsonnet runtime kind rather than exact JVM/Native class. This preserves the existing specialized string/number/array sort paths while making subtype-based string optimizations semantically transparent to the standard library.
Modification:
SetModule.getClasschecks in default sorting and key-function sorting with sort-kind checks.new_test_suite/set_sort_ascii_safe_strings.jsonnetfor all-AsciiSafeStr, mixedStr/AsciiSafeStr, and key-function-producedAsciiSafeStrsorting.Benchmark Results:
This is a correctness/enabler PR, not a claimed performance win. On the Scala Native 0.5.12 stacked profiling branch, kube-prometheus output stayed byte-identical; sort-kind-only A/B was neutral: forward
194.3 ± 14.6 msclean vs187.9 ± 12.3 mscandidate, reversed170.7 ± 2.7 msclean vs172.0 ± 1.6 mscandidate.Analysis:
The exact-class check was brittle because optimized runtime value subclasses must remain semantically indistinguishable from their base Jsonnet type. The replacement also avoids closure-based
forall(_.getClass == keyType)checks in this path, but the important result is that futureAsciiSafeStrpropagation can be evaluated without breakingstd.sort.References:
AsciiSafeStrpropagation for the Scala Native 0.5.12 performance work.Result:
./mill --no-server --ticker false --color false __.reformatpassed../mill --no-server --ticker false --color false -j 1 __.testpassed 444/444.