Parser executes trailing comment text on shebang lines
维护者通常 2 天内回复
还没有人认领这个 Issue。
评估
调研方向
Start with the t/testhelp shebang reproduction and compare its behavior across the JVM and interpreter backends. Run the focused command, then the relevant subprocess-response cases in t/gzip.t. Done means the complete shebang comment is discarded, @ARGV retains the three arguments, and the CGI-Compress-Gzip response tests pass without annotation or argument text.
由索引模型根据 Issue 内容生成。
描述
Summary
PerlOnJava does not fully discard a shebang/comment line when it contains a second # comment marker. Both backends execute text from the trailing Perl::Critic annotation as Perl code/output and consequently mishandle the script arguments that follow.
Reproduction
CGI-Compress-Gzip 1.03 contains t/testhelp, whose first line is:
#!perl -w ## no critic (RequireVersionVar)
Run the helper with its normal test arguments:
jperl -Iblib/arch -Iblib/lib t/testhelp -DHTTP_ACCEPT_ENCODING=gzip simple 'Hello World!'
Expected, as on standard Perl: a gzip response for Hello World!.
Actual on both the JVM and interpreter backends: the response is uncompressed, reports that the user agent does not accept gzip, and its body begins with:
nocritic(RequireVersionVar)-DHTTP_ACCEPT_ENCODING=gzipsimpleHello World!
The trailing comment contents and command-line arguments must not become program output. Perl should discard the complete shebang/comment line before compiling the script.
CPAN evidence
In CPAN random run 20260925-103132-20451, CGI-Compress-Gzip 1.03 failed 21 of 44 subtests in t/gzip.t on the JVM backend. The subprocess-response cases are directly caused by this parser defect. The same source passes all 44 tests under the standard-Perl oracle.
The focused helper reproduction also fails on the interpreter backend, so this is shared parser behavior rather than a JVM-only emission issue.
Acceptance criteria
- A script beginning with
#!perl -w ## no critic (RequireVersionVar)does not emit any part of that line. - Its
@ARGVremains-DHTTP_ACCEPT_ENCODING=gzip,simple, andHello World!when launched with those arguments. - The focused reproducer passes on standard Perl, PerlOnJava JVM, and PerlOnJava interpreter backends.
- CGI-Compress-Gzip's subprocess-response tests no longer contain the annotation or argument text in their bodies.
- 主要语言
- Perl
- 星标
- 64
- 派生
- 6
- 平均合并
- 5 小时
- 30 天内合并 PR
- 161
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
-
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 2 天内回复
-
area:backend area:cpan-port bug
难度 4/5 3-5 天 新手友好度 48/100
fglock/PerlOnJava#1548 ·
维护者通常 2 天内回复
-
bug
难度 5/5 一周以上 新手友好度 35/100
fglock/PerlOnJava#1541 ·
维护者通常 2 天内回复
-
area:backend area:cpan-port area:parser area:runtime bug
难度 3/5 1-2 天 新手友好度 78/100
fglock/PerlOnJava#1528 ·
维护者通常 2 天内回复
-
area:backend area:runtime bug
难度 4/5 3-5 天 新手友好度 65/100
fglock/PerlOnJava#1523 ·
维护者通常 2 天内回复
查看 fglock/PerlOnJava 的全部 Issue
相似的 Issue
-
难度 2/5 1-3 小时 新手友好度 84/100
维护者通常 1 天内回复
-
难度 1/5 1 小时以内 新手友好度 86/100
openfoodfacts/openfoodfacts-server#14724 ·
维护者通常 1 天内回复
-
难度 2/5 1-3 小时 新手友好度 72/100
rubocop/ruby-style-guide#982 ·
-
question
难度 2/5 1-3 小时 新手友好度 78/100
维护者通常 1 天内回复
-
难度 1/5 1 小时以内 新手友好度 75/100