Inline Traits & Specialized Traits#78
Closed
starswap wants to merge 649 commits into
Closed
Conversation
…a#25908) When a transitive dependency is missing from the classpath, the TASTy reader for a TYPEREF can produce a TypeRef whose symbol cannot be resolved. Downstream code in `computeBaseData` then hits an internal `non-class parent` assertion and crashes the compiler. Replace such unresolvable `TypeRef`s with a stub class symbol (similar to what Scala2Unpickler does), so the user sees a `BadSymbolicReference` error on first use instead of a compiler crash. The substitution is limited to references whose prefix is a package or a module, where member lookup is deterministic, to avoid converting transient lookup failures into spurious stubs during incremental completion. Adds an sbt scripted regression test (`sbt-test/tasty-compat/i20010`) that mirrors the original minimisation: three modules, `a`/`b`/`c`, where compiling `c` must not crash when `a` is absent from its classpath. Fixes scala#20010 ## How much have you relied on LLM-based tools in this contribution? Extensively, but its a small fix. ## How was the solution tested? New automated tests (including the issue's reproducer, if applicable)
…a#25980) `assumedCanEqual` lifts the scrutinee type by mapping non-opaque abstract types to their upper bound. When a GADT pattern object is matched against an invariant parameterized scrutinee (e.g. matching `object Bar extends Foo[Int]` against `Foo[A]`), this lifting raises `Foo[A]` to `Foo[Any]`, against which `Foo.Bar.type` is not a subtype because `Foo` is invariant — so the check fails even though GADT reasoning would correctly constrain `A := Int`. Try the unlifted subtype check first so GADT reasoning gets a chance, falling back to the existing lifted check. Fixes scala#25979 ## How much have you relied on LLM-based tools in this contribution? Extensively, but it's a simple fix. ## How was the solution tested? New automated tests (including the issue's reproducer, if applicable) --------- Co-authored-by: Claude <noreply@anthropic.com>
<!-- Fixes #XYZ (where XYZ is the issue number from the issue tracker) --> Implementation of scala#22298: Short description [JEP 409](https://openjdk.org/jeps/409) introduced sealed classes which in concept are similar to sealed trait, so in this PR we added a possibility to do exhaustiveness check in pattern matching for sealed abstract java classes. <!-- Ideally should have a title like "Fix #XYZ: Short fix description" --> <!-- TODO first sign the CLA https://contribute.akka.io/cla/scala --> <!-- if the PR is still a WIP, create it as a draft PR (or convert it into one) --> ## How much have you relied on LLM-based tools in this contribution? minimally <!-- State clearly in the pull request description, whether LLM-based tools were used and to what extent (extensively/moderately/minimally/not at all) --> <!-- Refer to our [LLM usage policy](https://github.com/scala/scala3/blob/main/LLM_POLICY.md) for rules and guidelines regarding usage of LLM-based tools in contributions. --> ## How was the solution tested? <!-- Tick exactly one of these boxes. Exceptions must have a very good reason. --> - [ ] Non-code change, no tests needed - [ ] Covered by existing tests (e.g., for refactorings) - [x] New automated tests (including the issue's reproducer, if applicable) - [ ] Manual tests described below (in enough detail that someone unfamiliar with this can still follow) Added automated test for getting a warn message if pattern matching is not exhaustive ## Additional notes <!-- Placeholder for any extra context regarding this contribution. --> <!-- When in doubt, and for support regarding contributions to a particular component of the compiler, refer to [our contribution guide](https://github.com/scala/scala3/blob/main/CONTRIBUTING.md), and feel free to tag the maintainers listed there for the area(s) you are modifying. --> --------- Co-authored-by: z.akhatuly <z.akhatuly@tbank.ru>
## Summary Fix REPL silently exiting after the first user input. The recent JLine refactor in scala#25925 added `reader.close()` to the `finally` block of `withMonitoringCtrlC`, called after every `interpret()`. But `terminal.reader()` returns the terminal's shared `NonBlockingReader`, so that close propagates to the terminal's input side. The next `readLine()` then gets immediate EOF and the REPL exits silently — typically right after the user's first command. ## Fix Drop the `reader.close()` call. The monitor thread's `reader.read(100L)` already returns within 100ms, and `userInput.signalClosed()` flips the loop condition so the monitor thread exits on its own. The `thread.join()` afterwards now waits up to ~100ms instead of returning immediately, which is negligible per command. This was fully Clauded. ## Test plan - [x] All 172 existing REPL tests pass (`sbt scala3-repl/test`) - [x] Manual: `bin/replQ` then multiple `val` definitions in sequence — REPL stays alive - [x] Manual: `import language.experimental.captureChecking` followed by `val x = new Object` — REPL stays alive and shows `val x: Object^{} = …` (capture checking active)
…(root, convert, generic)
Workaround for sbt/sbt#9185. As suggested in scala#25973 (comment).
Previously, compiling ```scala import scala.language.experimental.erasedDefinitions class Ev trait Svc[A]: def run(using erased Ev)(a: A): String class ISvc extends Svc[Int]: def run(using erased Ev)(a: Int): String = a.toString ``` crashed in erasure with `bad adapt for this.run(): (x$0: Int): String`. The bridge body is `Apply(this.run, [a])`. However, previously, `Erasure.Typer.typedApply` filters erased arguments using the function's *pre-erasure* type, and zip arguments against `paramErasureStatuses` positionally (`(erased Ev)(a: Int)`). Hence, `a` is mistaken for the erased slot and dropped, that leads to bad adaption. This commit fixes the issue, by just skipping the filter when the Apply is inside a bridge body. Fixes scala#25726.
When a class has a self-type that refines a member, the symbol of a TypeDef in its body can resolve through the self-type prefix to a different symbol than the one being defined. This caused two problems: - During TASTy unpickling, `tree.ensureHasSym(sym)` for a TypeDef tried to call `sym.asTerm.name`, which crashes with "asTerm called on not-a-Term" when sym is a type symbol. - During typer, the TypeDef tree's symbol was never reconciled with the intended sym, so `-Ycheck:all` would report the trait body as missing the type member. Use `sym.name` in `ensureHasSym` so the produced `NamedType` correctly selects between `TypeRef` and `TermRef` based on the name kind, and call `ensureHasSym` from `typedTypeDef` so the typed tree carries the right symbol before pickling. https://claude.ai/code/session_01Gu4rX4Ywk7zkf8yPqvvwNG
…#26192) Previously, they erroneously erased to `BoxedUnit`. Fixes #34653. ## How much have you relied on LLM-based tools in this contribution? Not at all ## How was the solution tested? New automated tests (including the issue's reproducer, if applicable) Note that this can break binary compatibility of libraries recompiled with a newer compiler. The issue contains a workaround: scala#24653 (comment) @WojciechMazur We had decided to include this in 3.9.0, but I was completely underwater until now. So this should be backported to 3.9.0, please. :)
Commit 9c971be made the Java parser consume annotations before array dimensions. That also permits annotations before a varargs ellipsis, since varargs parameters pass through the same type parsing path. Add a regression test for that syntax: ```java void varargs(String @lib.Valid ... values) {} ``` ## How much have you relied on LLM-based tools in this contribution? Minimally, for analysing the code. ## How was the solution tested? New test
Java permits type-use annotations before wildcard type arguments, for example ```java Function<@nullable ? super String, @nullable ? extends Object> ``` ## Reproduce You can use this repo's test files: ``` $ cd tests/pos/java-annotated-wildcards/ $ scala -version Scala code runner version: 1.12.4 Scala version (default): 3.8.3 $ scalac -usejavacp Nullable.java Test.java -- Error: Test.java:4:21 --------------------------------------------------------------------------------------------------------------------------- 4 | Function<@nullable ? super String, @nullable ? extends Object> mappingFunction(); | ^^^^^^^^ | illegal start of type -- Error: Test.java:4:82 --------------------------------------------------------------------------------------------------------------------------- 4 | Function<@nullable ? super String, @nullable ? extends Object> mappingFunction(); | ^ | null expected but ';' found. | 5 |} 2 errors found ``` Java however can compile it: ``` $ java -version openjdk version "1.8.0_492" OpenJDK Runtime Environment (Temurin)(build 1.8.0_492-b09) OpenJDK 64-Bit Server VM (Temurin)(build 25.492-b09, mixed mode) $ javac Nullable.java Test.java $ ls -1 *.class Nullable.class Test.class ``` ## How much have you relied on LLM-based tools in this contribution? <!-- Pick one; refer to https://github.com/scala/scala3/blob/main/LLM_POLICY.md for details --> Not at all (update: after the GitHub actions jobs failed I used codex for analysing the code and find out what broke, that lead to force push) ## How was the solution tested? New automated tests (including the issue's reproducer) Plus Manual tests ## See also - scala#22391 - basically same kind of fix
…6191) Fixes scala#26190. Copies erased flag into inline trait access proxies if the underlying field is erased. ## How much have you relied on LLM-based tools in this contribution? Minimally to search for things in the codebase. ## How was the solution tested? New automated tests.
_This PR applies 5/5 suggestions from code quality [AI findings](https://github.com/scala/scala3/security/quality/ai-findings)._ --------- Co-authored-by: Copilot Autofix powered by AI <223894421+github-code-quality[bot]@users.noreply.github.com>
_This PR applies 1/1 suggestions from code quality [AI findings](https://github.com/scala/scala3/security/quality/ai-findings)._ Co-authored-by: Copilot Autofix powered by AI <223894421+github-code-quality[bot]@users.noreply.github.com>
Fixes scala#25723 This seems to have been fixed in the meantime. ## How much have you relied on LLM-based tools in this contribution? Not at all ## How was the solution tested? New automated tests
…ialized to Any etc)
Add multiple param packs test
Author
|
Moved to scala#26156 |
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
20 participants
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.
Still DRAFT - running CI now
How much have you relied on LLM-based tools in this contribution?
How was the solution tested?
Additional notes