Prototype (&;@) calls consume following fat-comma chains

Open
#2,154 0 comments 0 reactions 0 assignees View on GitHub

Maintainers usually reply within 1 day

Nobody has claimed this yet.

Assessment

Difficulty
3/5
Estimated time
1-2 days
Newbie friendliness
45/100
Issue type
Bug
Clarity
Clearly specified
Activity status
Active
Tech stack
java, perl
Domain
compilers

Research direction

Start by reading PrototypeArgs.parsePrototypeArguments around the & handler and handleListOrHashArgument in src/main/java/org/perlonjava/frontend/parser/PrototypeArgs.java, then run the reproducer from the issue on both backends to see the failure near from Str. Trace how needComma propagates into the optional @ slurpy slot and decide when the list parse must stop at a => boundary. Done means the reproducer parses without parentheses, existing prototype behavior with real commas still works, and a new regression test is validated against system Perl and both PerlOnJava backends.

Written by the indexing model from the issue text.

Description

area:parser bug

Summary

PerlOnJava rejects a valid chained coercion declaration when an unparenthesized call with an ampersand-and-slurpy prototype is followed by a fat-comma chain. This prevents Types::XMLSchema 0.01 from loading on both JVM and interpreter backends.

Minimal reproducer

sub coerce { }
sub from (@) { }
sub via (&;@) { }

coerce X => from Int, via { 1 } => from Str, via { 2 };
print "ok\n";

System Perl accepts this and prints ok. PerlOnJava reports a syntax error near from Str. Parenthesizing the callback call as via({ 1 }) allows the corresponding chain to parse.

Root cause

Type::Utils::via has prototype (&;@). In PrototypeArgs.parsePrototypeArguments, the & handler parses a block and returns false for needComma. The following optional @ slot then calls handleListOrHashArgument. That method only requires a comma when needComma is true; with false, it parses a list starting at the outer => and consumes the next coercion clause as part of via's arguments. Parsing then fails at from Str.

The relevant code is in PrototypeArgs.java and handleListOrHashArgument. The parser source at the runtime's build commit is identical to the current source.

CPAN evidence

  • Distribution: Types-XMLSchema-0.01, run 20261119-133823-51077.
  • System Perl oracle passed all 34 tests. PerlOnJava failed t/00-compile.t, t/00-load.t, and t/types-xmlschema.t while loading Types/XMLSchema.pm at line 89.
  • The reduced reproducer fails on both JVM and interpreter backends; the equivalent parenthesized call parses on both.
  • This distribution is pure Perl; no XS dependency or reverse-dependency count applies.
  • Related issue #1744 covers a separate Type::Library extends failure on the interpreter and does not explain this parse error.

Environment and source identity

The PerlOnJava runtime reports Perl 5.44.1 and was built from commit 2de296a36d1b90e39241eab4c04fb2d93a179737; the current source checkout is 18742b1bf3acb794786a8dec51ec8f166aee6fb6, with the relevant parser files unchanged between them. The archived system-Perl oracle used Perl 5.42.2. The reproduction host is macOS arm64 with Java 24.0.2.

Acceptance criteria

  • Parse the reproducer above without parentheses on both backends.
  • Keep the optional slurpy prototype argument behavior correct when a comma actually introduces arguments.
  • Add project-owned regression coverage, validate its Perl semantics with system Perl, and run it on both PerlOnJava backends.
  • Confirm Types-XMLSchema-0.01 loads and its focused upstream tests pass after the fix.
Dominant language
Perl
Stars
67
Forks
8
Avg merge
5h 12m
Merged PRs (30d)
131

Getting set up

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from fglock/PerlOnJava

All issues in fglock/PerlOnJava

Similar issues

More Perl issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.