import json
import re
from datetime import datetime, timezone, timedelta
from typing import Any

TZ_CN = timezone(timedelta(hours=8))


def _now_collect_time() -> str:
    return datetime.now(TZ_CN).strftime("%Y-%m-%d %H:%M:%S")


def format_collected_at(value: Any = None) -> str:
    if value is None or (isinstance(value, str) and not value.strip()):
        return _now_collect_time()
    if isinstance(value, datetime):
        dt = value
        if dt.tzinfo is not None:
            dt = dt.astimezone(TZ_CN)
        return dt.strftime("%Y-%m-%d %H:%M:%S")
    s = str(value).strip()
    if re.fullmatch(r"\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}", s):
        return s
    if re.fullmatch(r"\d{4}-\d{2}-\d{2}", s):
        return f"{s} 00:00:00"
    try:
        dt = datetime.fromisoformat(s.replace("Z", "+00:00"))
        if dt.tzinfo is not None:
            dt = dt.astimezone(TZ_CN)
        return dt.strftime("%Y-%m-%d %H:%M:%S")
    except Exception:
        pass
    m = re.match(r"^(\d{4})[/-](\d{1,2})[/-](\d{1,2})[ T](\d{1,2}):(\d{1,2})(?::(\d{1,2}))?", s)
    if m:
        y, mo, d, h, mi, sec = m.groups()
        return f"{int(y):04d}-{int(mo):02d}-{int(d):02d} {int(h):02d}:{int(mi):02d}:{int(sec or 0):02d}"
    print(f"  [WARN] collectedAt 无法解析 {s!r}, 使用当前时间")
    return _now_collect_time()


def build_submit_payload(task: dict, parsed: dict, location: str = "") -> dict:
    result = parsed.get("result") or {}
    sources = parsed.get("sources") or []
    answer = (result.get("answer") or "").strip()
    think = (result.get("thinking_process") or "").strip() or None
    related = result.get("related_questions") or []
    if isinstance(related, list):
        extended = "\n".join(str(x) for x in related if x) or None
    else:
        extended = str(related) if related else None

    search_links = []
    for i, sr in enumerate(sources, 1):
        url = sr.get("source_url") or ""
        if not url:
            continue
        search_links.append({
            "url": url, "index": i,
            "title": sr.get("source_title") or "",
            "source": sr.get("source_name") or "",
            "snippet": sr.get("snippet") or sr.get("body") or None,
            "site_name": sr.get("source_name") or "",
            "site_icon": sr.get("source_icon") or "",
            "published_time": sr.get("source_publish_date") or "",
        })

    adopted = [{"url": u, "title": ""} for u in (result.get("cited_links") or []) if u]
    shared = []
    share_link = result.get("share_link") or ""
    if share_link:
        shared.append({"url": share_link, "title": ""})

    meta_infos = []
    for img in (result.get("image_links") or []):
        if img:
            meta_infos.append({
                "source_type": "image", "url": img, "title": "", "summary": "",
                "website_name": "", "icon": "",
                "value": json.dumps({"images": [{"full_size_url": [img]}], "videos": []}, ensure_ascii=False),
            })
    for vid in (result.get("video_links") or []):
        if vid:
            meta_infos.append({
                "source_type": "video", "url": vid, "title": "", "summary": "",
                "website_name": "", "icon": "",
                "value": json.dumps({"images": [], "videos": [{"url": vid}]}, ensure_ascii=False),
            })

    deep = int(task.get("deepThinking") if task.get("deepThinking") is not None
               else parsed.get("deep_thinking") or 0)
    web = int(task.get("webSearch") if task.get("webSearch") is not None
              else (1 if parsed.get("is_online") else 0))
    task_id = str(task.get("id") or "").strip()
    dispatch_id = str(task.get("dispatchId") or "").strip()
    collected_at = format_collected_at(
        parsed.get("collect_time") or (parsed.get("result") or {}).get("collect_time")
        or task.get("collectedAt")
    )

    return {
        "id": task_id, "dispatchId": dispatch_id, "collectedAt": collected_at,
        "answer": answer, "thinkText": think,
        "deepThinking": 1 if deep else 0, "webSearch": 1 if web else 0,
        "model": task.get("model") or "", "location": location or "",
        "extendedQuestion": extended, "searchLinks": search_links,
        "adoptedLinks": adopted, "sharedLinks": shared,
        "metaInfos": meta_infos, "productLinkList": [],
    }