Moo role composition loses lexical-sub closures
维护者通常 1 天内回复
还没有人认领这个 Issue。
评估
调研方向
The issue is in the JVM backend's handling of lexical closures during Moo role composition. Start by examining the minimal reproducer in the issue body to understand the failure. Look at how PerlOnJava implements subroutine aliasing and environment preservation, particularly in the Moo::Role composition path. The fix must ensure lexical pads are retained when methods are copied. Run the provided CPAN test for CXC::Form::Tiny::Plugin::OptArgs2 to verify the resolution.
由索引模型根据 Issue 内容生成。
描述
Summary
PerlOnJava loses the lexical-sub closure when a Moo role copies a method into a consuming class. This breaks Form::Tiny plugin metaclasses and causes CXC::Form::Tiny::Plugin::OptArgs2 0.12 to fail during class construction.
Minimal reproducer
The following program prints 42 under system Perl:
package R; use Moo::Role; my sub helper { 41 }; sub role_method { helper()+1 }; package C; use Moo; with "R"; package main; print C->new->role_method, "\\n";
On the JVM backend it prints the expected 42 but also emits:
Subroutine "&helper" is not available
The warning is emitted from the role-composed method after it has been installed in C. This demonstrates that the method body retained its call to the lexical helper but lost the lexical environment that should contain helper. System Perl exits successfully without the warning.
CPAN evidence
The CPAN regression run 20260924-085657-82687 reported CXC::Form::Tiny::Plugin::OptArgs2 0.12 failing in t/OptArgs.t, t/inherit.t, t/nested.t, and t/required.t. The failure is:
Can.t locate object method "new" via package "My::Form"
The same run emitted unavailable-lexical-sub diagnostics for CXC::Form::Tiny::Plugin::OptArgs2::Meta methods referencing croak, _match_inherit_optargs, and parse_inherit_matches. t/optargs_opts.t passed.
With the exact dependency blibs and load order, the five substantive upstream test files pass under system Perl: 15 tests, all successful. Therefore this is attributable to PerlOnJava rather than upstream behavior or missing prerequisites.
Backend results
- JVM backend: minimal reproducer emits the unavailable lexical-sub diagnostic; the CPAN test reaches My::Form->new and fails.
- Interpreter backend: the same dependency stack currently fails earlier while loading the Moo/Role::Tiny implementation with a strict-refs error involving Form::Tiny::Error::with. It does not invalidate the JVM reproducer; interpreter parity should be checked separately.
Likely ownership
The primary defect is in closure/environment preservation when role composition aliases or installs a subroutine into another package. A method containing a lexical-sub call must retain the defining lexical pad after Moo::Role composition. The downstream missing constructor is a consequence of the Form::Tiny metaclass/class role setup failing after these composed methods cannot resolve their lexical helpers.
Acceptance criteria
- The minimal Moo role reproducer passes on the JVM backend without an unavailable-lexical-sub diagnostic.
- CXC::Form::Tiny::Plugin::OptArgs2 0.12 passes its five substantive test files on the JVM backend.
- The interpreter backend has an explicit parity result for the same lexical-role-closure case.
- Add a permanent project-owned regression test; verify it with system Perl first and on both PerlOnJava backends after the fix.
No upstream tests were modified.
- 主要语言
- Perl
- 星标
- 64
- 派生
- 7
- 平均合并
- 5 小时 25 分钟
- 30 天内合并 PR
- 213
环境准备
- 提供 Dockerfile 或 Docker Compose 文件
- 没有 Pull Request 模板
- 阅读贡献指南
从这里开始
- 先读完整个 Issue，再读项目的贡献指南。
- 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
- Fork 仓库，在一个分支上完成修改。
- 提交 Pull Request，并在描述里引用这个 Issue 编号。
fglock/PerlOnJava 的其他 Issue
-
难度 2/5 1-3 小时 新手友好度 68/100
fglock/PerlOnJava#1651 ·
维护者通常 1 天内回复
-
bug
难度 2/5 1-3 小时 新手友好度 72/100
fglock/PerlOnJava#1579 ·
维护者通常 1 天内回复
-
area:cpan-port area:unicode bug
难度 2/5 1-3 小时 新手友好度 88/100
fglock/PerlOnJava#1341 ·
维护者通常 1 天内回复
-
难度 3/5 半天 新手友好度 45/100
fglock/PerlOnJava#1808 ·
维护者通常 1 天内回复
-
area:cpan-port area:runtime bug
难度 4/5 3-5 天 新手友好度 28/100
fglock/PerlOnJava#1802 · 1 条评论 ·
维护者通常 1 天内回复
查看 fglock/PerlOnJava 的全部 Issue
相似的 Issue
-
Add .gitattributes so share/skel keeps LF on a Windows git checkout可能已有人在做 @sadhana-s-codez 于 1 天前认领。 未关闭
难度 1/5 1 小时以内 新手友好度 85/100
PerlDancer/Dancer2#1854 · 1 条评论 ·
维护者通常 1 天内回复
-
bug
难度 2/5 1-3 小时 新手友好度 78/100
major/MySQLTuner-perl#1007 ·
维护者通常 1 天内回复
-
难度 1/5 1 小时以内 新手友好度 88/100
-
难度 1/5 1 小时以内 新手友好度 88/100
-
难度 2/5 1-3 小时 新手友好度 69/100
bioconda/bioconda-recipes#70187 ·
维护者通常 2 天内回复