#!/usr/bin/env python3
"""Halo 全量导出（Phase 2 T1）：posts/photos/moments → .migration/*.json

对生产仅 GET；正文优先取列表 status.content，缺则回落 head-content 端点。
产物为迁移中间件，不入库（.gitignore .migration/）。
"""
import json
import sys
import urllib.request
from pathlib import Path

BASE = "http://127.0.0.1:18090"
OUT = Path(".migration")
OUT.mkdir(exist_ok=True)


def get(path: str) -> object:
    with urllib.request.urlopen(BASE + path, timeout=30) as r:
        return json.load(r)


def export_list(api: str, size: int = 100) -> list[dict]:
    """翻页取全量 items。"""
    items, page = [], 1
    while True:
        data = get(f"{api}?page={page}&size={size}")
        batch = data.get("items", [])
        items.extend(batch)
        if page * size >= data.get("total", 0) or not batch:
            return items
        page += 1


def main() -> int:
    posts = export_list("/apis/api.content.halo.run/v1alpha1/posts")
    # 正文兜底：head-content 拿完整 HTML（status.content 可能是截断摘要）
    for p in posts:
        name = p["metadata"]["name"]
        content = (p.get("status") or {}).get("content")
        if not content:
            try:
                head = get(f"/apis/api.content.halo.run/v1alpha1/posts/{name}/head-content")
                p["_fullContent"] = head.get("content", "")
            except Exception as e:  # noqa: BLE001 - 单篇失败不阻断全量导出
                p["_fullContent"] = ""
                print(f"  warn: {name} head-content 失败: {e}", file=sys.stderr)
    photos = export_list("/apis/api.photo.halo.run/v1alpha1/photos")
    try:
        groups = get("/apis/api.photo.halo.run/v1alpha1/groups")
    except Exception:  # noqa: BLE001 - 分组端点缺失时空兜底
        groups = {"items": []}
    moments = export_list("/apis/api.moment.halo.run/v1alpha1/moments")

    (OUT / "posts.json").write_text(json.dumps(posts, ensure_ascii=False, indent=1))
    (OUT / "photos.json").write_text(json.dumps(photos, ensure_ascii=False, indent=1))
    (OUT / "photo_groups.json").write_text(json.dumps(groups, ensure_ascii=False, indent=1))
    (OUT / "moments.json").write_text(json.dumps(moments, ensure_ascii=False, indent=1))

    full = sum(1 for p in posts if p.get("_fullContent") or (p.get("status") or {}).get("content"))
    print(f"exported: posts={len(posts)}（含正文 {full}）photos={len(photos)} groups={len(groups.get('items', []))} moments={len(moments)}")
    return 0


if __name__ == "__main__":
    sys.exit(main())