life-echo/api/app/features/memoir/chapter_markdown_compose.py

"""
按 chapter_story_links 顺序将各 story 正文物化为单一 markdown（无 LLM）。
保留 story 内 asset:// 引用不变。
章节级 canonical：仅正文拼接，故事间用 ---；故事标题仅存 stories.title。
PDF 导出可单独物化「## 标题 + 正文」版本。
"""

from typing import Any

from app.features.memoir.markdown_sanitize import sanitize_story_for_chapter_compose


def _gather_title_body_pairs(chapter: Any) -> list[tuple[str, str]]:
    links = sorted(
        list(getattr(chapter, "story_links", None) or []),
        key=lambda x: getattr(x, "order_index", 0),
    )
    pairs: list[tuple[str, str]] = []
    for link in links:
        st = getattr(link, "story", None)
        if st is None:
            continue
        title = (getattr(st, "title", None) or "").strip()
        body = (getattr(st, "canonical_markdown", None) or "").strip()
        pairs.append((title, body))
    return pairs


def compose_ordered_stories_to_markdown(ordered: list[tuple[str, str]]) -> str:
    """
    :param ordered: (story_title, canonical_markdown) 已按阅读顺序排好（title 仅用于清洗去重）
    :return: 章节级 markdown；仅各故事正文，非空块之间用 \\n\\n---\\n\\n 分隔
    """
    bodies: list[str] = []
    for title, md in ordered:
        raw = (md or "").strip()
        if not raw:
            continue
        cleaned = sanitize_story_for_chapter_compose(raw, title)
        if cleaned:
            bodies.append(cleaned)
    return "\n\n---\n\n".join(bodies)


def compose_ordered_stories_to_pdf_markdown(ordered: list[tuple[str, str]]) -> str:
    """PDF：每故事 ## 标题 + 正文，块间 ---（标题来自元数据，不写回章节 canonical）。"""
    parts: list[str] = []
    for title, md in ordered:
        t = (title or "").strip() or "故事"
        raw = (md or "").strip()
        if not raw:
            continue
        body = sanitize_story_for_chapter_compose(raw, title)
        if not body:
            continue
        parts.append(f"## {t}\n\n{body}")
    return "\n\n---\n\n".join(parts)


def materialize_chapter_markdown_from_loaded_chapter(chapter: Any) -> str:
    """要求 chapter.story_links 已 eager-load，且各 link.story 可用。"""
    return compose_ordered_stories_to_markdown(_gather_title_body_pairs(chapter))


def materialize_chapter_pdf_markdown_from_loaded_chapter(chapter: Any) -> str:
    """PDF 专用：含每段 ## 故事名。"""
    return compose_ordered_stories_to_pdf_markdown(_gather_title_body_pairs(chapter))