Triage: potential duplicate open PRs

[toddr-bot]

Potential Duplicate Pull Requests

A review of the 112 open PRs identified the following clusters where PRs appear to address the same problem, modify the same files, or would conflict with each other. Consolidating these would reduce review burden and prevent merge conflicts.

1. UTF-8 decoding when use utf8 is present (EXACT DUPLICATE)

• PR fix: Decode UTF-8 source when 'use utf8' pragma is present #333 — "Decode UTF-8 source when 'use utf8' pragma is present"
• PR fix: decode UTF-8 source when 'use utf8' pragma is present #363 — "decode UTF-8 source when 'use utf8' pragma is present"

Near-identical titles, same files changed (Tokenizer.pm, t/14_charsets.t), same approach (utf8::decode() in Tokenizer::new()). The only difference is which issue each references. One should be closed in favor of the other.

2. Encoding support for PPI::Document->new

• PR fix: add encoding parameter to PPI::Document->new #438 — adds an encoding parameter to Document->new()
• PR feat: add encoding-aware constructor APIs #434 — adds named constructors (new_from_string, new_from_file, new_from_handle) with encoding support

Both add an encoding() accessor and modify Document.pm. They solve the same problem with different API designs. At most one approach should be adopted.

3. Whitespace between sigil and identifier

• PR allow variables with spaces after the sigil #307 — "allow variables with spaces after the sigil"
• PR fix: allow whitespace between sigil and identifier #388 — "allow whitespace between sigil and identifier"

Both modify Token/Symbol.pm and Token/Unknown.pm. Same problem, likely same fix.

4. cmp_document functions

• PR cmp_document branch #49 — "cmp_document branch"
• PR Cmp document functions #146 — "Cmp document functions"

Both add document comparison. PR #146 is a much larger changeset that appears to subsume #49.

5. Minimum Perl version bump / IO::String removal

• PR fix: bump minimum Perl to 5.008, remove IO::String fallback #362 — bump minimum Perl to 5.008, remove IO::String fallback
• PR fix: bump minimum perl to 5.008001 (Lancaster Consensus) #393 — bump minimum perl to 5.008001 (Lancaster Consensus)
• PR fix: remove unnecessary IO::String prereq declaration #386 — remove unnecessary IO::String prereq declaration

All three modify dist.ini, Token/Data.pm, and test files. PR #393 (5.008001) supersedes #362 (5.008). PR #386's IO::String removal is a subset of #362's changes.

6. PPI::Exception caller tracking

• PR fix: populate callers on object throw with no prior callers #390 — "populate callers on object throw with no prior callers"
• PR fix: add PPI::Exception unit tests and fix instance throw() caller tracking #391 — "add PPI::Exception unit tests and fix instance throw() caller tracking"

Both fix the same bug in Exception.pm (instance throw() not recording callers). #391 adds more comprehensive tests. One subsumes the other.

7. Element->namespace() and package block handling

• PR fix: handle scoped namespaces via Element->namespace #399 — "handle scoped namespaces via Element->namespace"
• PR fix: implement package block(), _complete(), and Element namespace() #414 — "implement package block(), _complete(), and Element namespace()"
• PR fix: _complete() for block-form package declarations #376 — "_complete() for block-form package declarations"

All three modify Statement/Package.pm. PRs #399 and #414 both implement Element->namespace() and Package->block(). PR #414 appears to be the most complete, likely subsuming both #399 and #376.

8. DESTROY cleanup safety

• PR fix: skip DESTROY cleanup during global destruction #353 — "skip DESTROY cleanup during global destruction"
• PR fix: stop Node::DESTROY from corrupting still-referenced children #404 — "stop Node::DESTROY from corrupting still-referenced children"

Both modify Node.pm and address DESTROY-related corruption. Different approaches to the same root problem — will likely conflict.

9. __DATA__ / __END__ / Pod parsing

• PR fix: parse POD within __DATA__ sections #349 — "parse POD within DATA sections"
• PR fix: parse Pod below __END__ when __DATA__ section present #439 — "parse Pod below END when DATA section present"
• PR fix: __DATA__ no longer causes incomplete documents #377 — "DATA no longer causes incomplete documents"

All three address how the tokenizer handles __DATA__ sections. PRs #349 and #439 both modify Token/Data.pm and deal with Pod extraction from data sections. These will conflict with each other.

10. Extensibility / Plugin mechanisms

• PR feat: add PPI::Plugin system for parse-time extensibility #437 — "add PPI::Plugin system for parse-time extensibility"
• PR feat: add register_statement_class API for extending PPI grammar #360 — "add register_statement_class API for extending PPI grammar"
• PR feat: add custom_keywords for runtime keyword extension #373 — "add custom_keywords for runtime keyword extension"

All three modify Lexer.pm and add different extension mechanisms. These are competing designs for the same need — at most one should be adopted, or they need to be reconciled into a unified approach.

11. insert_before / insert_after improvements

• PR Fix up PPI::Token::insert_{after,before}. #79 — "Fix up PPI::Token::insert_{after,before}"
• PR Allow statements as well as structures and tokens to be inserted adjacent to a token #151 — "Allow statements as well as structures and tokens to be inserted adjacent to a token"
• PR fix: clarify insert_before/insert_after docs for single element #367 — "clarify insert_before/insert_after docs for single element"
• PR fix: Allow insert_before/insert_after to accept multiple elements, strings, and fragments #392 — "Allow insert_before/insert_after to accept multiple elements, strings, and fragments"
• PR docs: cross-link insert method rules and clarify PDOM explanation #419 — "cross-link insert method rules and clarify PDOM explanation"

Five PRs touching the same insert API surface. PR #151 is a massive changeset. PR #392 extends the API. PRs #367 and #419 are docs-only. These need coordinated review — merging any one will likely break the others.

---

Other related (but probably distinct) PRs worth noting

These groups touch related areas but appear to solve different sub-problems. Flagging them in case they conflict at the code level:

• Filename/Document metadata: PR actual filename and filename attribute #182 (filename attribute) and PR fix: document filename param for PPI::Document->new #379 (document filename param) — both modify Document.pm and t/29_logical_filename.t
• Location/offset tracking: PR Add byte offsets for Tokens #186 (byte offsets), PR fix: add document character offset to element locations #415 (character offset), PR fix: auto-invalidate location cache after document mutations #420 (cache invalidation), PR fix: preserve location data through clone operations #423 (clone preservation) — four PRs touching element location data
• Feature scoping in Lexer: PR Fix feature scoping leak when leaving blocks #351 (scoping leak fix), PR factor out the feature set stack from the lexer #357 (factor out feature stack), PR fix: handle whitespace tokens in feature argument decomposition #335 (whitespace in feature args) — all touch the Lexer's feature tracking
• Unicode/non-Latin parsing: PR fix: parse Unicode letters in unquoted hash keys #331 (Unicode hash keys) and PR fix: handle non-latin variable names in tokenizer #369 (non-latin variable names) — different tokenizer areas but related UTF-8 concern that overlaps with cluster 1
• Quote method consistency: PR fix: add PPI::Token::QuoteLike->string, like PPI::Token::Quote has #406 (QuoteLike->string) and PR fix: make methods on PPI::Token::Quote subclasses consistent #411 (Quote subclass methods) — different class hierarchies but same API consistency goal

---

🤖 Generated by Kōan from PR audit session