feat(migrate): Halo 全量内容迁移(压缩版)——6文/21图/1瞬间 URL 1:1

- 正文源=线上渲染页(Halo 快照层 patch 残迹:泥潭/赎罪仅 2 字符),
  pandoc html→gfm + tags API 显示名换算后分类:5 书评(五档情绪色)+1 技术
- 图库 21 张:长边 2400/q86 压缩入仓(原图 196M→9.5M——公开仓历史膨胀
  与 nginx 100M 推送上限双杀;原图在 Phase 0 备份 tar 中永久留档)
- 瞬间 1 条;URL 1:1(6 slug 全等,零 301)
- 本提交由 7 个本地提交重写而成(清除已进历史的 196M blob,未推送过);
  因 reset --soft 暂存竞态,一并包含 spec 归档文件
  (.spec/specs/archive/2026-10-24_migrate-blog-content/)

验证: check_urls.py 6==6;astro build 11 页 exit 0(8.67s);
      EXIF 管线跑通;staging 更新;halo healthy+200 全程未动

Co-authored-by: GLM-5.3
This commit is contained in:
wait4xandGLM-5.3 committed 2026-10-25 00:39:08 +08:00
1 parent 7b71acf223
commit aea21c8694
74 files changed
+1381 -97

No files matched your search

+52
View File
@@ -0,0 +1,52 @@
#!/usr/bin/env python3
"""瞬间迁移(Phase 2 T4):Halo moments → content/moments/*.md
Halo moment.content.html 为富文本(内联样式很脏),经 pandoc 提纯为
纯文本 md;medium(配图)当前库存为空,字段留待将来有图时扩展。
"""
import json
import subprocess
from datetime import datetime, timezone
from pathlib import Path
OUT = Path("content/moments")
def main() -> int:
moments = json.load(open(".migration/moments.json", encoding="utf-8"))
n = 0
for m in moments:
spec = m["spec"]
if spec.get("visible") == "PRIVATE":
continue
html = (spec.get("content") or {}).get("html", "")
text = subprocess.run(
["/home/wait4x/.local/bin/pandoc", "-f", "html", "-t", "plain", "--wrap=none"],
input=html.encode(), capture_output=True, check=True,
).stdout.decode("utf-8").strip()
if not text:
continue
ts = spec.get("releaseTime") or m["metadata"].get("creationTimestamp", "")
day = datetime.fromisoformat(ts.replace("Z", "+00:00")).astimezone(timezone.utc).strftime("%Y-%m-%d")
out = OUT / f"{day}.md"
i = 2
while out.exists():
out = OUT / f"{day}-{i}.md"
i += 1
tags = spec.get("tags") or []
# schema 约束 text/image 在 frontmatter(body 不参与校验),一句话即 text 字段
out.write_text(
"---\n"
f"pubDate: {day}\n"
f"tags: [{', '.join(tags)}]\n"
f'text: "{text.replace(chr(34), chr(39))}"\n'
"---\n",
encoding="utf-8",
)
n += 1
print(f"migrated {n} moments")
return 0
if __name__ == "__main__":
raise SystemExit(main())