Package stash iteration hides named subs from runtime traversal
維護者通常 1 天內回覆
還沒有人認領這個 Issue。
評估
研究方向
The bug is in PerlOnJava's stash iteration, so start with the code that implements %{*{"::"}} and per-package stash hashes, checking whether named subroutine CODE slots are exposed. Reproduce with the four-line Perl script in the issue, expecting found=Bar1::bar. Done means that script prints the expected output and a focused regression test for enumerating a declared package sub passes on both JVM and interpreter backends.
由索引模型根據 Issue 內容生成。
描述
Root cause
Aspect 1.04 discovers statically matched subs in Aspect::Pointcut::match_all by iterating package stashes (each %{*{"::"}} and then each package stash) and checking defined *ENTRY{CODE}. PerlOnJava's stash iteration does not expose named subroutine entries through that traversal: with Bar1::bar declared, Aspect::Pointcut::Call->new('Bar1::bar')->match_all prints Bar1::bar on system Perl but an empty list on PerlOnJava. As a result, Aspect installs no advice wrapper for the target; before, after, and around advice never run. The same behavior occurs on both JVM and interpreter backends.
Reproducer
use strict;
use Aspect;
package Bar1;
sub bar { 'value' }
package main;
my $p = Aspect::Pointcut::Call->new('Bar1::bar');
print 'found=', join(',', $p->match_all), "\n";
Expected: found=Bar1::bar
Actual on PerlOnJava: found=
Evidence
- Distribution: Aspect 1.04, CPAN run
20261008-181412-1954. - System Perl 5.42.2: full upstream suite passed, 28 files / 873 tests.
- PerlOnJava 5.44.1, JVM and interpreter: full run failed 22/28 files, 168/671 subtests; focused
t/31_feature_caller.tfailed the nine expected advice-count assertions on both backends. - Root cause confirmed directly by comparing
match_alloutput for the reproducer on system Perl and PerlOnJava. - Environment: macOS arm64, OpenJDK 24.0.2.
Fix package-stash iteration so named CODE entries are visible to Aspect's runtime stash traversal. Add focused regression coverage for enumerating a declared package sub through this traversal, then rerun the Aspect suite on both backends.
- 主要語言
- Perl
- 星號
- 64
- 分支
- 7
- 平均合併
- 5 小時 25 分鐘
- 30 天內合併 PR
- 178
環境準備
- 提供 Dockerfile 或 Docker Compose 檔案
- 沒有 Pull Request 範本
- 閱讀貢獻指南
從這裡開始
- 先讀完整個 Issue，再讀專案的貢獻指南。
- 在 Issue 下留言說明你要接手 —— 這能避免兩個人做同樣的事。
- Fork 儲存庫，在一個分支上完成修改。
- 送出 Pull Request，並在描述裡引用這個 Issue 編號。
fglock/PerlOnJava 的其他 Issue
-
難度 2/5 1-3 小時 新手友好度 68/100
fglock/PerlOnJava#1651 ·
維護者通常 1 天內回覆
-
bug
難度 2/5 1-3 小時 新手友好度 72/100
fglock/PerlOnJava#1579 ·
維護者通常 1 天內回覆
-
area:cpan-port area:unicode bug
難度 2/5 1-3 小時 新手友好度 88/100
fglock/PerlOnJava#1341 ·
維護者通常 1 天內回覆
-
難度 5/5 一週以上 新手友好度 12/100
fglock/PerlOnJava#1696 ·
維護者通常 1 天內回覆
-
documentation enhancement
難度 3/5 半天 新手友好度 50/100
fglock/PerlOnJava#1694 ·
維護者通常 1 天內回覆
查看 fglock/PerlOnJava 的全部 Issue
相似的 Issue
-
bug help wanted
難度 2/5 1 小時以內 新手友好度 70/100
bioepic-data/bervo#145 · 1 則留言 ·
維護者通常 1 天內回覆
-
難度 2/5 1-3 小時 新手友好度 68/100
-
Help-Wanted Needs-Triage Package-Update
難度 2/5 1-3 小時 新手友好度 72/100
microsoft/winget-pkgs#448814 · 1 則留言 ·
維護者通常 1 天內回覆
-
FOODTURE
難度 2/5 1-3 小時 新手友好度 67/100
openfoodfacts/openfoodfacts-server#14853 · 1 則留言 ·
維護者通常 1 天內回覆
-
難度 1/5 1 小時以內 新手友好度 85/100
gugod/App-perlbrew#879 ·