Commit Graph

14 Commits

Author SHA1 Message Date
d052f9dc97 feat: using構文完全実装&json_native大幅進化
## 🎉 using構文の完全実装(ChatGPT作業)
-  **include → using移行完了**: 全ファイルでusing構文に統一
  - `local X = include` → `using "path" as X`
  - 約70ファイルを一括変換
-  **AST/パーサー/MIR完全対応**: using専用処理実装
  - ASTNode::Using追加
  - MIRビルダーでの解決処理
  - include互換性も維持

## 🚀 json_native実装進化(ChatGPT追加実装)
-  **浮動小数点対応追加**: is_float/parse_float実装
-  **配列/オブジェクトパーサー実装**: parse_array/parse_object完成
-  **エスケープ処理強化**: Unicode対応、全制御文字サポート
-  **StringUtils大幅拡張**: 文字列操作メソッド多数追加
  - contains, index_of_string, split, join等
  - 大文字小文字変換(全アルファベット対応)

## 💡 MIR SIMD & ハイブリッド戦略考察
- **MIR15 SIMD命令案**: SimdLoad/SimdScan等の新命令セット
- **C ABIハイブリッド**: ホットパスのみC委託で10倍速化可能
- **並行処理でyyjson超え**: 100KB以上で2-10倍速の可能性
- **3層アーキテクチャ**: Nyash層/MIR層/C ABI層の美しい分離

## 📊 技術的成果
- using構文により名前空間管理が明確化
- json_nativeが実用レベルに接近(完成度25%→40%)
- 将来的にyyjsonの70%速度達成可能と判明

ChatGPT爆速実装×Claude深い考察の完璧な協働!

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-09-25 00:41:56 +09:00
5e818eeb7e stage3: unify to cleanup; MIR return-defer; docs+smokes updated; LLVM(harness): finalize_phis ownership, ret.py simplified, uses-predeclare; cleanup return override green; method-postfix cleanup return WIP (PHI head) 2025-09-19 02:07:38 +09:00
951a050592 selfhost: introduce using-based imports for compiler/parser/tools; keep includes temporarily. llvm: add PHI wiring JSON trace + unit/integration tests; fast test suite extended. runner: split selfhost helpers, small cleanups. 2025-09-18 13:35:38 +09:00
3fe908eb0d docs/runtime: externcall + println normalization; NyRT silent result flag; PyVM console warn/error/trace routing\nselfhost(M2): add MirEmitterBox (Return(Int)->const+ret) and --emit-mir switch; initial selfhost smoke\nrunner: child args/env hardening for selfhost (NYASH_NY_COMPILER_CHILD_ARGS under --; NYASH_JSON_ONLY=1, NYASH_VM_USE_PY=1, NYASH_DISABLE_PLUGINS=1); VM path strips using lines when NYASH_ENABLE_USING=1\nmeta: clarify PyVM/llvmlite as primary (vm/interpreter legacy OFF) in CURRENT_TASK.md; README links to externcall docs 2025-09-18 06:35:49 +09:00
9a7d77d868 selfhost parser: add // and /* */ comment skipping; support \n\r\t and \uXXXX escapes in strings (read_string_lit/parse_string2); add tools/selfhost_parser_json_smoke.sh (optional) 2025-09-18 04:38:14 +09:00
2720884a20 bridge/json_v0: split expr lowering; add Ternary/Peek lowering + AST; selfhost Peek JSON emit; add selfhost Peek smoke; warning cleanup in lowering/optimizer/verification
- Split expr lowering into ; route calls from stmt lowering
- Implement ternary/peek lowering (MIR13 PHI-off=Copy, PHI-on=Phi)
- Extend JSON v0 AST (ExprV0::{Ternary,Peek}, PeekArmV0)
- Selfhost parser_box: emit Peek JSON; add Stage-2 'Peek basic' smoke
- Reduce warnings: remove unused imports/vars in several modules
- current_task: update plan for legacy VM/Interpreter offboarding
2025-09-17 11:45:57 +09:00
adbfbb2c76 ci: add GitHub Actions workflow; runner: prefer PyVM in selfhost paths; stage3: add LLVM + bridge acceptance smokes; docs: update env flags 2025-09-17 01:20:15 +09:00
97a76c0571 selfhost: stub stage3 parser nodes under gate 2025-09-16 18:36:54 +09:00
fa1619bf4b selfhost: add ternary parser and plugin prefix guard 2025-09-16 17:38:22 +09:00
7db2c24808 selfhost: harden parser Stage-2 defaults 2025-09-16 17:10:27 +09:00
63c8fda808 🔍 Research: GPT-5-Codex capabilities and GitHub PR integration
## Summary
Investigated OpenAI's new GPT-5-Codex model and Codex GitHub PR review integration capabilities.

## GPT-5-Codex Analysis

### Benchmark Performance (Good)
- SWE-bench Verified: 74.5% (vs GPT-5's 72.8%)
- Refactoring tasks: 51.3% (vs GPT-5's 33.9%)
- Code review: Higher developer ratings

### Real-World Issues (Concerning)
- Users report degraded coding performance
- Scripts that previously worked now fail
- Less consistent than GPT-4.5
- Longer response times (minutes vs instant)
- "Creatively and emotionally flat"
- Basic errors (e.g., counting letters incorrectly)

### Key Finding
Classic case of "optimizing for benchmarks vs real usability" - scores well on tests but performs poorly in practice.

## Codex GitHub PR Integration

### Setup Process
1. Enable MFA and connect GitHub account
2. Authorize Codex GitHub app for repos
3. Enable "Code review" in repository settings

### Usage Methods
- **Manual**: Comment '@codex review' in PR
- **Automatic**: Triggers when PR moves from draft to ready

### Current Limitations
- One-way communication (doesn't respond to review comments)
- Prefers creating new PRs over updating existing ones
- Better for single-pass reviews than iterative feedback

## 'codex resume' Feature
New session management capability:
- Resume previous codex exec sessions
- Useful for continuing long tasks across days
- Maintains context from interrupted work

🐱 The investigation reveals that while GPT-5-Codex shows benchmark improvements, practical developer experience has declined - a reminder that metrics don't always reflect real-world utility\!
2025-09-16 16:28:25 +09:00
94d95dfbcd 🚀 Break/Continue/Try-Catch構文のサポート追加とMIRループ制御強化
## 主な変更点

### 🎯 MIRループ制御の実装(根治対応)
- src/mir/loop_builder.rs: Break/Continue対応のループコンテキスト管理
  - ループのbreak/continueターゲットブロック追跡
  - ネストループの適切な処理
- src/mir/builder.rs: Break/Continue文のMIR生成実装
- src/tokenizer.rs: Break/Continue/Tryトークン認識追加

### 📝 セルフホストパーサーの拡張
- apps/selfhost-compiler/boxes/parser_box.nyash:
  - Stage-3: break/continue構文受理(no-op実装)
  - Stage-3: try-catch-finally構文受理(構文解析のみ)
  - エラー処理構文の将来対応準備

### 📚 ドキュメント更新
- 論文K(爆速事件簿): 45事例に更新(4件追加)
  - PyVM迂回路の混乱事件
  - Break/Continue無限ループ事件
  - EXE-first戦略の再発見
- 論文I(開発秘話): Day 38の重要決定追加

### 🧪 テストケース追加
- apps/tests/: ループ制御とPHIのテストケース
  - nested_loop_inner_break_isolated.nyash
  - nested_loop_inner_continue_isolated.nyash
  - loop_phi_one_sided.nyash
  - shortcircuit関連テスト

## 技術的詳細
- Break/ContinueをMIRレベルで適切に処理
- 無限ループ問題(CPU 99.9%暴走)の根本解決
- 将来の例外処理機能への準備

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-09-15 22:14:42 +09:00
d90216e9c4 📚 Phase 15 - セルフホスティング戦略の明確化とEXE-first実装
## 主な変更点

### 🎯 戦略の転換と明確化
- PyVMを開発ツールとして位置づけ(本番経路ではない)
- EXE-first戦略を明確に優先(build_compiler_exe.sh実装済み)
- Phase順序の整理: 15.2(LLVM)→15.3(コンパイラ)→15.4(VM)

### 🚀 セルフホスティング基盤の実装
- apps/selfhost-compiler/にNyashコンパイラMVP実装
  - compiler.nyash: メインエントリー(位置引数対応)
  - boxes/: parser_box, emitter_box, debug_box分離
- tools/build_compiler_exe.sh: ネイティブEXEビルド+dist配布
- Python MVPパーサーStage-2完成(local/if/loop/call/method/new)

### 📝 ドキュメント整備
- Phase 15 README/ROADMAP更新(Self-Hosting優先明記)
- docs/guides/exe-first-wsl.md: WSLクイックスタート追加
- docs/private/papers/: 論文G~L、爆速事件簿41事例収録

### 🔧 技術的改善
- JSON v0 Bridge: If/Loop PHI生成実装(ChatGPT協力)
- PyVM/llvmliteパリティ検証スイート追加
- using/namespace機能(gated実装、Phase 15では非解決)

## 次のステップ
1. パーサー無限ループ修正(未実装関数の実装)
2. EXEビルドとセルフホスティング実証
3. c0→c1→c1'ブートストラップループ確立

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-09-15 18:44:49 +09:00
af11c6855b 🚀 Start Phase 15.3: Nyash compiler MVP implementation
Major milestone:
- Set up apps/selfhost-compiler/ directory structure
- Implement basic Nyash compiler in Nyash (CompilerBox)
- Stage-1: Basic arithmetic parser (int/string/+/-/*/括弧/return)
- JSON v0 output compatible with --ny-parser-pipe
- Runner integration with NYASH_USE_NY_COMPILER=1 flag
- Comprehensive smoke tests for PHI/Bridge/Stage-2

Technical updates:
- Updated CLAUDE.md with Phase 15.3 status and MIR14 details
- Statement separation policy: newline-based with minimal ASI
- Fixed runaway ny-parser-pipe processes (CPU 94.9%)
- Clarified MIR14 as canonical instruction set (not 13/18)
- LoopForm strategy: PHI auto-generation during reverse lowering

Collaborative development:
- ChatGPT5 implementing compiler skeleton
- Codex provided LoopForm PHI generation guidance
- Claude maintaining documentation and coordination

🎉 セルフホスティングの歴史的一歩!自分自身をコンパイルする日が近いにゃ!

Co-Authored-By: ChatGPT <noreply@openai.com>
2025-09-15 01:21:37 +09:00