208 lines
6.5 KiB
Markdown
208 lines
6.5 KiB
Markdown
---
|
||
phase: 13
|
||
plan: 06
|
||
type: execute
|
||
wave: 2
|
||
depends_on:
|
||
- 13-01
|
||
- 13-02
|
||
- 13-03
|
||
- 13-04
|
||
- 13-05
|
||
files_modified:
|
||
- kb_workflow/srt/semantic_engine.py
|
||
- tests/test_semantic_engine.py
|
||
autonomous: true
|
||
requirements:
|
||
- STRAT-01
|
||
- STRAT-02
|
||
- STRAT-03
|
||
must_haves:
|
||
truths:
|
||
- "系统提供可扩展的分段策略架构"
|
||
- "Phase 13 实现统一策略（用户决策：无类型区分）"
|
||
- "策略架构支持未来添加新策略（扩展性）"
|
||
artifacts:
|
||
- path: "kb_workflow/srt/semantic_engine.py"
|
||
provides: "可扩展的策略架构（Strategy Pattern 预留）"
|
||
min_lines: 300
|
||
exports: ["SemanticSegmentationEngine", "SegmentationStrategy"]
|
||
- path: "tests/test_semantic_engine.py"
|
||
provides: "策略扩展性测试"
|
||
min_lines: 50
|
||
key_links:
|
||
- from: "SegmentationStrategy"
|
||
to: "SemanticSegmentationEngine"
|
||
via: "策略模式架构"
|
||
pattern: "class.*Strategy.*segment"
|
||
- from: "tests"
|
||
to: "SegmentationStrategy"
|
||
via: "验证策略可扩展性"
|
||
pattern: "test.*strategy.*extensibility"
|
||
---
|
||
|
||
<objective>
|
||
实现可扩展的分段策略架构
|
||
|
||
目的：提供策略架构支持未来扩展（Phase 13 实现统一策略，遵守用户决策）
|
||
输出：SegmentationStrategy 接口，统一策略实现
|
||
</objective>
|
||
|
||
<execution_context>
|
||
@/home/starmen1265/.claude/get-shit-done/workflows/execute-plan.md
|
||
@/home/starmen1265/.claude/get-shit-done/templates/summary.md
|
||
</execution_context>
|
||
|
||
<context>
|
||
@.planning/phases/13-llm/13-CONTEXT.md
|
||
@.planning/phases/13-llm/13-RESEARCH.md
|
||
|
||
# User Decision (Locked)
|
||
- **视频类型策略**: 无类型区分 — 使用统一的分段 Prompt，不针对不同视频类型（知识类/故事类/教程类）设计不同策略
|
||
|
||
# Research Document
|
||
Open Question 3: STRAT-01/STRAT-02/STRAT-03 需求的边界
|
||
Recommendation: 在代码结构上预留策略扩展接口（Strategy Pattern），但 Phase 13 只实现统一策略
|
||
</context>
|
||
|
||
<feature>
|
||
<name>分段策略架构</name>
|
||
<files>kb_workflow/srt/semantic_engine.py, tests/test_semantic_engine.py</files>
|
||
<behavior>
|
||
架构设计（研究文档建议）：
|
||
- 预留 Strategy Pattern 接口（SegmentationStrategy）
|
||
- Phase 13 实现统一策略（UnifiedSegmentationStrategy）
|
||
- 支持未来添加知识类/故事类等特定策略
|
||
|
||
Phase 13 实现（遵守用户决策）：
|
||
- 统一策略适用于所有视频类型
|
||
- 视频类型仅用于 Prompt 上下文，不改变分段策略
|
||
|
||
测试验证：
|
||
- 策略接口可扩展
|
||
- 统一策略适用于所有视频类型
|
||
</behavior>
|
||
<implementation>
|
||
在 semantic_engine.py 中：
|
||
1. 定义 SegmentationStrategy Protocol（接口）
|
||
2. 实现 UnifiedSegmentationStrategy（统一策略）
|
||
3. SemanticSegmentationEngine 使用策略接口
|
||
|
||
测试：
|
||
1. test_strategy_selection() - 验证策略选择
|
||
2. test_strategy_extensibility() - 验证策略可扩展性
|
||
</implementation>
|
||
</feature>
|
||
|
||
<tasks>
|
||
<task type="auto">
|
||
<name>Task 1: 实现可扩展策略架构</name>
|
||
<files>kb_workflow/srt/semantic_engine.py, tests/test_semantic_engine.py</files>
|
||
<action>
|
||
在 semantic_engine.py 中实现策略架构：
|
||
|
||
1. 定义 SegmentationStrategy Protocol：
|
||
```python
|
||
from typing import Protocol, List
|
||
from kb_workflow.srt.parser import Segment
|
||
|
||
class SegmentationStrategy(Protocol):
|
||
"""分段策略接口"""
|
||
|
||
def segment(
|
||
self,
|
||
segments: List[Segment],
|
||
metadata: dict
|
||
) -> List[str]:
|
||
"""对字幕分段进行语义分段
|
||
|
||
Args:
|
||
segments: 字幕分段列表
|
||
metadata: 视频元数据
|
||
|
||
Returns:
|
||
段落文本列表
|
||
"""
|
||
...
|
||
```
|
||
|
||
2. 实现 UnifiedSegmentationStrategy（统一策略）：
|
||
```python
|
||
class UnifiedSegmentationStrategy:
|
||
"""统一分段策略
|
||
|
||
适用于所有视频类型（知识类、故事类、访谈类等）。
|
||
视频类型仅用于 Prompt 上下文，不改变分段策略。
|
||
|
||
遵守用户决策：无类型区分，使用统一 Prompt。
|
||
"""
|
||
|
||
def __init__(self, ai_client: Optional[AIClientProtocol] = None):
|
||
self.ai_client = ai_client
|
||
# Lazy initialization...
|
||
|
||
def segment(
|
||
self,
|
||
segments: List[Segment],
|
||
metadata: dict
|
||
) -> List[str]:
|
||
# 调用 LLM 进行语义分段
|
||
# 使用统一的 SEGMENTATION_PROMPT
|
||
...
|
||
```
|
||
|
||
3. 更新 SemanticSegmentationEngine：
|
||
```python
|
||
class SemanticSegmentationEngine:
|
||
def __init__(
|
||
self,
|
||
strategy: Optional[SegmentationStrategy] = None,
|
||
fallback_segmenter: Optional['SemanticSegmenter'] = None
|
||
):
|
||
# 默认使用统一策略
|
||
self.strategy = strategy or UnifiedSegmentationStrategy()
|
||
self.fallback_segmenter = fallback_segmenter or SemanticSegmenter()
|
||
|
||
def segment(
|
||
self,
|
||
segments: List[Segment],
|
||
metadata: dict
|
||
) -> List[str]:
|
||
return self.strategy.segment(segments, metadata)
|
||
```
|
||
|
||
测试验证：
|
||
- 策略接口可扩展（可实现新策略）
|
||
- 统一策略适用于所有视频类型
|
||
- 策略切换不影响功能
|
||
</action>
|
||
<verify>
|
||
<automated>pytest tests/test_semantic_engine.py::test_strategy_extensibility -x</automated>
|
||
</verify>
|
||
<done>
|
||
- SegmentationStrategy Protocol 定义完成
|
||
- UnifiedSegmentationStrategy 实现完成
|
||
- SemanticSegmentationEngine 使用策略接口
|
||
- 测试验证策略可扩展性
|
||
</done>
|
||
</task>
|
||
</tasks>
|
||
|
||
<verification>
|
||
- pytest tests/test_semantic_engine.py -v 相关测试通过
|
||
- 策略接口清晰（Protocol 定义）
|
||
- 统一策略符合用户决策（无类型区分）
|
||
- 架构支持未来扩展
|
||
</verification>
|
||
|
||
<success_criteria>
|
||
1. SegmentationStrategy Protocol 定义完成
|
||
2. UnifiedSegmentationStrategy 实现完成
|
||
3. SemanticSegmentationEngine 使用策略接口
|
||
4. 测试验证策略可扩展性
|
||
</success_criteria>
|
||
|
||
<output>
|
||
After completion, create `.planning/phases/13-llm/13-06-SUMMARY.md`
|
||
</output>
|