116 lines
3.2 KiB
Python
116 lines
3.2 KiB
Python
|
"""
|
|||
|
|
从标题/正文推断 Story 的人生时间键，供章节内按发生顺序排序。
|
|||
|
|
"""
|
|||
|
|
|
|||
|
|
from __future__ import annotations
|
|||
|
|
|
|||
|
|
import re
|
|||
|
|
from datetime import datetime
|
|||
|
|
from typing import Any
|
|||
|
|
|
|||
|
|
_YEAR_IN_TITLE = re.compile(r"(?:^|[\s·，,])(?P<y>(?:19|20)\d{2})(?:年|\s|·|$)")
|
|||
|
|
_YEAR_ANYWHERE = re.compile(r"(?:19|20)\d{2}")
|
|||
|
|
|
|||
|
|
|
|||
|
|
def parse_year_from_title(title: str | None) -> int | None:
|
|||
|
|
"""从「1999年 · 标题」或标题中的四位年份提取年份。"""
|
|||
|
|
if not title:
|
|||
|
|
return None
|
|||
|
|
m = _YEAR_IN_TITLE.search(title.strip())
|
|||
|
|
if m:
|
|||
|
|
try:
|
|||
|
|
return int(m.group("y"))
|
|||
|
|
except ValueError:
|
|||
|
|
return None
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
|
|||
|
|
def parse_year_from_text(text: str | None, *, max_chars: int = 4000) -> int | None:
|
|||
|
|
"""从正文前若干字中取第一个 plausible 四位年份。"""
|
|||
|
|
if not text:
|
|||
|
|
return None
|
|||
|
|
chunk = (text or "")[:max_chars]
|
|||
|
|
m = _YEAR_ANYWHERE.search(chunk)
|
|||
|
|
if m:
|
|||
|
|
try:
|
|||
|
|
y = int(m.group(0))
|
|||
|
|
if 1900 <= y <= 2100:
|
|||
|
|
return y
|
|||
|
|
except (ValueError, TypeError):
|
|||
|
|
return None
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
|
|||
|
|
def parse_time_start_year(time_start: str | None) -> int | None:
|
|||
|
|
"""time_start 存 'YYYY' 或 'YYYY-MM' 等，取年份用于排序。"""
|
|||
|
|
if not time_start or not str(time_start).strip():
|
|||
|
|
return None
|
|||
|
|
s = str(time_start).strip()
|
|||
|
|
m = re.match(r"^(\d{4})", s)
|
|||
|
|
if m:
|
|||
|
|
try:
|
|||
|
|
y = int(m.group(1))
|
|||
|
|
if 1900 <= y <= 2100:
|
|||
|
|
return y
|
|||
|
|
except ValueError:
|
|||
|
|
return None
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
|
|||
|
|
def apply_infer_story_time_start_to_model(story: Any) -> None:
|
|||
|
|
"""将推断的 YYYY 写入 `story.time_start`（已有合法值则保留）。"""
|
|||
|
|
ts = infer_story_time_start(
|
|||
|
|
title=getattr(story, "title", None),
|
|||
|
|
canonical_markdown=getattr(story, "canonical_markdown", None),
|
|||
|
|
existing_time_start=getattr(story, "time_start", None),
|
|||
|
|
)
|
|||
|
|
if ts:
|
|||
|
|
story.time_start = ts
|
|||
|
|
|
|||
|
|
|
|||
|
|
def infer_story_time_start(
|
|||
|
|
*,
|
|||
|
|
title: str | None,
|
|||
|
|
canonical_markdown: str | None,
|
|||
|
|
existing_time_start: str | None = None,
|
|||
|
|
) -> str | None:
|
|||
|
|
"""
|
|||
|
|
返回统一 `YYYY` 字符串；若无法推断则返回 None。
|
|||
|
|
已有 time_start 合法则保留。
|
|||
|
|
"""
|
|||
|
|
y_existing = parse_time_start_year(existing_time_start)
|
|||
|
|
if y_existing is not None:
|
|||
|
|
return str(y_existing)
|
|||
|
|
|
|||
|
|
y_title = parse_year_from_title(title)
|
|||
|
|
if y_title is not None:
|
|||
|
|
return str(y_title)
|
|||
|
|
|
|||
|
|
y_body = parse_year_from_text(canonical_markdown)
|
|||
|
|
if y_body is not None:
|
|||
|
|
return str(y_body)
|
|||
|
|
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
|
|||
|
|
def life_sort_key_parts(
|
|||
|
|
*,
|
|||
|
|
time_start: str | None,
|
|||
|
|
title: str | None,
|
|||
|
|
created_at: datetime | None,
|
|||
|
|
story_id: str,
|
|||
|
|
) -> tuple[int, int, str]:
|
|||
|
|
"""
|
|||
|
|
用于 sorted(...): 人生发生顺序（早→晚）。
|
|||
|
|
无年份时排在后面（9999），再以创建时间、id 稳定 tie-break。
|
|||
|
|
"""
|
|||
|
|
y = parse_time_start_year(time_start)
|
|||
|
|
if y is None:
|
|||
|
|
y = parse_year_from_title(title)
|
|||
|
|
if y is None:
|
|||
|
|
y = 9999
|
|||
|
|
sub = 0
|
|||
|
|
if created_at is not None:
|
|||
|
|
sub = int(created_at.timestamp())
|
|||
|
|
return (y, sub, story_id)
|