ai-unit-test-generator 1.3.0

package/CHANGELOG.md

@@ -0,0 +1,264 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [1.0.0] - 2025-01-10
+
+### Breaking Changes
+- **Package renamed**: `ut-priority-scorer` → `ai-test-generator`
+- **CLI renamed**: `ut-score` → `ai-test`
+- Better reflects the package's core value: AI-powered test generation
+
+### Added
+- **AI Test Generation Tools**: Complete workflow for batch test generation
+  - `ai-test gen-prompt`: Generate AI prompts with function source code context
+  - `ai-test extract-tests`: Extract and create test files from AI responses  
+  - `ai-test mark-done`: Mark functions as DONE/SKIP for progress tracking
+- **Progress Tracking**: Status column (TODO/DONE/SKIP) in all reports
+- **Batch Generation**: Support for pagination with `--skip` option for large codebases
+- **Flexible Framework Support**: Customizable framework hints in prompts via `--framework` option
+- **Multiple AI Formats**: Test extraction supports various AI response formats
+
+### Changed
+- Reports now sorted by score (highest priority first) for better UX
+- Enhanced CLI with comprehensive help messages and examples
+- Improved documentation with end-to-end AI test generation workflow
+- Better error messages and validation across all commands
+
+### Fixed
+- Test extraction regex now supports multiple response formats (with/without backticks)
+- Improved function source code extraction for various code patterns
+- Better error handling in all CLI commands
+
+## Previous Versions (as ut-priority-scorer)
+
+### [1.0.3] - 2025-10-10
+- Sort output by total score (descending) for better prioritization
+- Added `Status` column (TODO/DONE/SKIP) for tracking test completion
+- Added Summary section with priority distribution statistics
+- Improved Markdown output with AI-friendly format
+
+### [1.0.2] - 2025-10-10
+- Fixed missing `esprima` dependency
+
+### [1.0.1] - 2025-10-10
+- Fixed missing `escomplex` dependency
+
+### [1.0.0] - 2025-10-10 (Initial Release as ut-priority-scorer)
+
+### Added
+- Initial release
+- Layered architecture scoring system (Foundation, Business Logic, State Management, UI)
+- CLI commands: `init`, `scan`
+- Flexible JSON configuration with sensible defaults
+- Comprehensive scoring metrics:
+  - Testability (pure function detection, mock complexity)
+  - Dependency count (module reference analysis)
+  - Complexity (cyclomatic + cognitive via ESLint)
+  - Business criticality (keyword-based detection)
+  - Error risk (Git history analysis)
+- Report generation in Markdown and CSV formats
+- Priority levels: P0 (≥7.5), P1 (6.0-7.4), P2 (4.0-5.9), P3 (<4.0)
+- AST-based target scanning using ts-morph
+- Git signal analysis using simple-git
+- ESLint + sonarjs integration for cognitive complexity
+- Dependency graph analysis for module references
+- Programmatic API for workflow integration
+- Comprehensive documentation and examples
+
+### Features
+- 🎯 AI-optimized scoring weights for test generation efficiency
+- 📊 Layered priority system aligned with testing pyramid principles
+- 🔧 Flexible JSON-based configuration
+- 🚀 Fast scanning and scoring (handles 300+ targets in seconds)
+- 📈 Rich reporting with detailed metrics breakdown
+- 🤖 Designed for AI-powered test generation workflows
+
+### Technical Details
+- Node.js ≥18.0.0 required
+- ES Module format
+- Zero runtime dependencies for core functionality
+- Minimal npm package size (~50KB)
+
+## [1.1.0] - 2025-01-11
+
+### Added
+- **Coverage Delta Validation**: `run-batch` now validates coverage improvement (P0: ≥2%, P1: ≥1%)
+- **Enhanced AI Output Parsing**: `extract-tests` supports 5+ response formats with fuzzy matching
+- **Stability Reruns**: P0 targets automatically rerun Jest once for flakiness detection
+- **Failure Hints Loop**: Jest failures auto-generate actionable hints for next batch
+
+### Changed
+- **Prompt Constraints**: Added "禁止修改源码/新增依赖/真实时间网络" stability guidelines
+- **Reports**: Coverage delta displayed in run-batch output (before/after/delta)
+- **Extract Robustness**: Support markdown variations (中/英文标题、有/无反引号、宽松路径匹配)
+
+### Inspired By
+- Meta TestGen-LLM: assurance filters (build/stability/coverage) - https://arxiv.org/abs/2402.09171
+- Airbnb: large-scale batching strategies
+- TestART: iterative generation and template-based repair - https://arxiv.org/abs/2408.03095
+- Midscene.js: stability practices - https://github.com/web-infra-dev/midscene
+
+### Technical
+- Default assurance mode: P0 strong (reruns=1, minCovDelta=2), P1 light, P2/P3 build-only
+- AI-agnostic: works with any LLM (cursor-agent/GPT/Claude/local models)
+- Zero breaking changes to existing commands
+
+## [1.2.0] - 2025-01-11
+
+### 🎉 Major Refactoring - Modular Architecture
+
+**Breaking Changes**
+- **File Structure**: Reorganized `lib/` into layered modules (core/ai/testing/workflows/utils)
+- **File Naming**: Renamed all lib files for clarity and consistency
+  - `gen-targets.mjs` → `core/scanner.mjs`
+  - `gen-git-signals.mjs` → `core/git-analyzer.mjs`
+  - `score-ut.mjs` → `core/scorer.mjs`
+  - `gen-test-prompt.mjs` → `ai/prompt-builder.mjs`
+  - `ai-generate.mjs` → `ai/client.mjs`
+  - `extract-tests.mjs` → `ai/extractor.mjs`
+  - `jest-runner.mjs` → `testing/runner.mjs`
+  - `jest-failure-analyzer.mjs` → `testing/analyzer.mjs`
+  - `run-batch.mjs` → `workflows/batch.mjs`
+  - `run-all.mjs` → `workflows/all.mjs`
+  - `mark-done.mjs` → `utils/marker.mjs`
+- **CLI**: Renamed `bin/ai-test.js` → `bin/cli.js` (internal, no user impact)
+
+### Added
+- **Modular Architecture**: Clear separation of concerns
+  - `lib/core/`: Code analysis & scoring (no AI dependency)
+  - `lib/ai/`: AI interaction & test generation
+  - `lib/testing/`: Test execution & failure analysis
+  - `lib/workflows/`: End-to-end orchestration
+  - `lib/utils/`: Shared utilities
+- **Module Exports**: Each layer has an `index.mjs` with clean exports
+- **Programmatic API**: Package now supports programmatic usage
+  ```javascript
+  import { scanCode, scoreTargets } from 'ai-test-generator'
+  import { buildPrompt, callAI } from 'ai-test-generator/ai'
+  import { runBatch } from 'ai-test-generator/workflows'
+  ```
+- **Package Exports**: Added `exports` field in package.json for subpath imports
+
+### Changed
+- **package.json**: Updated `main` to `lib/index.mjs`, `bin` to `cli.js`, version to 1.2.0
+- **Internal Paths**: All import paths updated to reflect new structure
+
+### Non-Breaking
+- ✅ CLI commands remain unchanged (`ai-test scan`, `ai-test gen-prompt`, etc.)
+- ✅ Config file format unchanged
+- ✅ Output formats unchanged
+- ✅ User workflow unchanged
+
+### Benefits
+- 🎯 **Clear Responsibility**: Each module has a single, well-defined purpose
+- 📦 **Easier Extension**: Add new analyzers, AI clients, or test runners without touching other code
+- 🧩 **Reusable Components**: Use scoring engine without AI, or AI layer without workflows
+- 📖 **Better Discoverability**: Flat structure → layered structure improves navigation
+- 🚀 **Future-Proof**: Prepared for Vitest support, OpenAI client, plugin system
+
+## [1.3.0] - 2025-01-11
+
+### 🎉 Major Simplification - 极简 CLI
+
+**Breaking Changes**
+- **CLI Commands**: 简化从 7 个命令到 2 个核心命令
+  - ✅ 保留: `scan`, ~~`init`~~ (自动化)
+  - ✅ 新增: `generate` (合并 `gen-prompt` + `extract-tests` + `run-batch` + `mark-done`)
+  - ❌ 移除: `init`, `gen-prompt`, `extract-tests`, `mark-done`, `run-batch`, `run-all`
+
+### Added
+- **`ai-test generate`**: 一键生成测试
+  - 默认生成 10 个 P0 TODO 函数
+  - 选项: `-n` 数量, `-p` 优先级, `--all` 全部生成
+  - 自动工作流: 读取 TODO → 生成 → 运行 Jest → 标记 DONE
+  - 智能重试: 失败的函数保持 TODO，下次自动带 hints
+  
+- **状态自动管理**:
+  - 测试通过 → 自动标记 DONE
+  - 测试失败 → 保持 TODO (带失败 hints)
+  - `scan` 重新运行时保留 DONE 状态
+  
+- **增量支持**:
+  - 新扫描保留已有 DONE/SKIP 状态
+  - 已删除的函数自动从报告移除
+  - 新函数自动加入 TODO
+
+### Changed
+- **`ai-test scan`**: 自动初始化配置（首次运行时）
+- **`ai-test scan`**: 显示 TODO/DONE 统计
+- **状态持久化**: DONE/SKIP 状态在重新扫描后保留
+- **`workflows/batch.mjs`**: 
+  - 只处理 TODO 状态的函数
+  - 成功后自动标记 DONE
+  - 失败保持 TODO 并生成 hints
+- **`core/scorer.mjs`**: 
+  - 读取旧报告的状态
+  - 保留 DONE/SKIP 到新报告
+
+### Removed
+- ❌ `ai-test init` - 自动化（scan 时创建）
+- ❌ `ai-test gen-prompt` - 内部使用
+- ❌ `ai-test extract-tests` - 内部使用
+- ❌ `ai-test mark-done` - 自动管理
+- ❌ `ai-test run-batch` - 合并到 generate
+- ❌ `ai-test run-all` - 合并到 generate --all
+
+### Migration Guide
+
+**v1.2.0 → v1.3.0**
+
+旧工作流:
+```bash
+ai-test init
+ai-test scan
+ai-test run-batch P0 10 0
+ai-test mark-done "funcA,funcB"
+```
+
+新工作流:
+```bash
+ai-test scan                  # 自动初始化
+ai-test generate              # 自动生成 + 标记
+```
+
+完整示例:
+```bash
+# 首次使用
+ai-test scan                  # 扫描代码
+ai-test generate              # 生成 10 个 P0
+ai-test generate -n 20        # 再生成 20 个
+ai-test generate --all        # 跑完所有 TODO
+
+# 增量更新
+# (添加新代码后)
+ai-test scan                  # 重新扫描（保留 DONE）
+ai-test generate              # 继续生成新的 TODO
+```
+
+### Benefits
+- 🎯 **更简单**: 2 个命令代替 7 个
+- 🤖 **全自动**: 无需手动标记状态
+- 🔄 **智能重试**: 失败自动带 hints
+- 📊 **进度可见**: 实时显示 TODO/DONE
+- ⚡ **增量友好**: 保留已完成的工作
+
+## [Unreleased]
+
+### Planned Features
+- VS Code extension for inline priority indicators
+- Watch mode for continuous scoring during development
+- HTML report with interactive charts
+- Plugin system for custom metrics
+- Integration with popular CI/CD platforms
+- Test coverage correlation analysis
+- AI test generation templates
+- Multi-language support (currently TypeScript/JavaScript only)
+
+---
+
+For a complete list of changes, see the [commit history](https://github.com/temptrip/ai-test-generator/commits).
+

package/LICENSE

@@ -0,0 +1,22 @@
+MIT License
+
+Copyright (c) 2025 YuhengZhou
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+