7f57f96c25
本次 squash merge 将 codex-story-first-image-intent 的整体改动合入 development，核心内容包括： 1. 后端数据与迁移：新增 stories、story_versions、story_image_intents、chapter_cover_intents、assets 等模型与 Alembic 迁移，建立 story-first、markdown-first、asset-first 的主数据链路。 2. 生成与任务链：引入 StoryBuilderOrchestrator、ChapterComposerOrchestrator、story_image_tasks、chapter_cover_tasks，图片生成从正文占位符改为结构化 intent -> asset -> markdown 回填。 3. 并发与一致性：为 story/chapter intent 增加 claim_token、claimed_at、attempt_count，采用数据库原子 claim 为主、Redis 锁为辅，避免重复生成、锁误删和 processing 卡死。 4. Memoir 读写路径：章节 canonical_markdown 成为正文真源，列表/详情接口补齐 markdown、cover_asset、word_count 等字段，PDF 与 asset 解析链路同步升级。 5. Memory / Retrieval：扩展 transcript ingest、chunking、evidence 检索与 story 聚合基础设施，为后续 story-first RAG 与多 agent 编排提供底座。 6. App 端体验：章节页继续走 MarkdownRenderer 阅读链，同时吸收 fix3-19 的跨平台 UI glitch 修复；更新对话页、首页、文案资源与章节列表映射逻辑。 7. 测试与文档：补充 asset resolver、story image task、章节封面派发、markdown 映射等回归测试，并加入图片占位符退役设计文档。
107 lines
3.4 KiB
Python
107 lines
3.4 KiB
Python
"""
|
||
ChapterComposerOrchestrator — 读取 stories/evidence，生成章节 markdown。
|
||
|
||
Agent 只产出结构化结果，不直接写 DB。
|
||
"""
|
||
|
||
from __future__ import annotations
|
||
|
||
from typing import Any
|
||
|
||
from app.core.logging import get_logger
|
||
|
||
logger = get_logger(__name__)
|
||
|
||
|
||
class ChapterComposerOrchestrator:
|
||
"""
|
||
生成章节大纲和章节 markdown。
|
||
仅返回 markdown，不落库。
|
||
"""
|
||
|
||
def compose_chapter_markdown(
|
||
self,
|
||
*,
|
||
title: str,
|
||
category: str,
|
||
evidence: dict,
|
||
existing_markdown: str = "",
|
||
user_profile: str = "",
|
||
birth_year: int | None = None,
|
||
llm: Any = None,
|
||
) -> str:
|
||
"""
|
||
从 evidence 生成章节 markdown。
|
||
若有 existing_markdown 则追加/合并。
|
||
返回 markdown 正文，不写 DB。
|
||
"""
|
||
from app.agents.memoir.narrative_agent import NarrativeAgent
|
||
|
||
chunks = evidence.get("relevant_chunks", [])
|
||
facts = evidence.get("relevant_facts", [])
|
||
new_content = self._format_evidence_for_prompt(chunks, facts)
|
||
|
||
agent = NarrativeAgent()
|
||
narrative = agent.generate_narrative(
|
||
stage=category,
|
||
slots={},
|
||
new_content=new_content,
|
||
existing_content=existing_markdown,
|
||
user_profile=user_profile,
|
||
birth_year=birth_year,
|
||
llm=llm,
|
||
)
|
||
return self._to_markdown(narrative)
|
||
|
||
def _format_evidence_for_prompt(self, chunks: list, facts: list) -> str:
|
||
"""将 evidence 格式化为 prompt 输入。"""
|
||
parts = []
|
||
for c in chunks[:10]:
|
||
content = (
|
||
c.get("content", "")
|
||
if isinstance(c, dict)
|
||
else getattr(c, "content", "")
|
||
)
|
||
if content:
|
||
parts.append(content.strip())
|
||
for f in facts[:5]:
|
||
if isinstance(f, dict):
|
||
subj = f.get("subject", "")
|
||
pred = f.get("predicate", "")
|
||
obj = f.get("object_json", "")
|
||
if subj or pred:
|
||
parts.append(f"{subj} {pred} {obj}")
|
||
else:
|
||
parts.append(
|
||
f"{getattr(f, 'subject', '')} {getattr(f, 'predicate', '')}"
|
||
)
|
||
return "\n\n".join(parts) if parts else ""
|
||
|
||
def _to_markdown(self, narrative: str) -> str:
|
||
"""将 narrative（JSON 或纯文本）转为 markdown。正文不含占位符。"""
|
||
if not narrative or not narrative.strip():
|
||
return ""
|
||
if narrative.strip().startswith("{") and "paragraphs" in narrative:
|
||
import json
|
||
|
||
try:
|
||
data = json.loads(narrative)
|
||
paras = data.get("paragraphs", [])
|
||
if isinstance(paras, list):
|
||
parts = []
|
||
for p in paras:
|
||
if isinstance(p, dict):
|
||
text = p.get("content", p.get("text", ""))
|
||
else:
|
||
text = str(p)
|
||
if text.strip():
|
||
parts.append(text.strip())
|
||
md = "\n\n".join(parts)
|
||
else:
|
||
md = narrative
|
||
except json.JSONDecodeError:
|
||
md = narrative
|
||
else:
|
||
md = narrative.strip()
|
||
return md
|