#!/usr/bin/env python3
"""X Account Monitor - Scan Chunk 2 (auto-discovered accounts)"""
import urllib.request, json, os, time
from datetime import datetime, timezone, timedelta

API_KEY = "get-x-api-6157f8edb420c3d5e50e062287389aa55635cd7c9630c81f"
BASE = "https://api.getxapi.com"
RESULTS_FILE = "/home/kiro/trend-engine/trends/x/account-monitor/scan_results_2026-06-13.json"
CUTOFF = datetime(2026, 6, 10, 0, 0, 0, tzinfo=timezone.utc)
SCAN_DATE = "2026-06-13 06:00 UTC"

# Relevance filter set
IRRELEVANT = {
    "CINEMA505", "TheAIColony", "GoogleLabs", "GoogleAI",
    "Ai_Vaidehi", "AI_with_jasmin", "CharaspowerAI",
    "AIBuzzNews", "AlexStudio44", "FaithJaiye2187",
    "KammelaBala", "MyStageAI",
}

HANDLES = [
    "AITOPIAai", "FaithJaiye2187", "KammelaBala", "MyStageAI",
    "AI_Video_Dumps", "ChrisLaubAI", "GoogleLabs", "TheAIColony", "cladlabs",
    "AIBuzzNews", "AlexStudio44", "DonutAI", "GoogleAI",
    "AI_with_jasmin", "CharaspowerAI",
    "Ai_Vaidehi",
    "CINEMA505",
    "AIFrontliner", "AIGuideHQ",
    "AIWarper",
    "AITECHio",
    "4aibsc",
]


def parse_date(created_at):
    if not created_at:
        return None
    try:
        return datetime.strptime(created_at, "%a %b %d %H:%M:%S %z %Y")
    except:
        return None


def fetch_tweets(handle):
    url = f"{BASE}/twitter/user/tweets?userName={handle}&count=5"
    req = urllib.request.Request(url, headers={"Authorization": f"Bearer {API_KEY}"})
    try:
        with urllib.request.urlopen(req, timeout=30) as resp:
            body = resp.read().decode("utf-8", errors="replace")
        data = json.loads(body, strict=False)
    except Exception as e:
        return {"error": str(e), "tweets": []}
    
    tweets = data.get("tweets", data) if isinstance(data, dict) else data
    if not isinstance(tweets, list):
        return {"error": f"unexpected response type: {type(data)}", "tweets": []}
    return {"tweets": tweets}


print(f"=== Scan Chunk 2 — {SCAN_DATE} ===")
print(f"Handles: {len(HANDLES)}")
print(f"Cutoff: {CUTOFF.isoformat()}")
print()

# Load existing results
all_results = {"accounts": [], "errors": [], "scan_date": SCAN_DATE}
if os.path.exists(RESULTS_FILE):
    with open(RESULTS_FILE) as f:
        all_results = json.load(f)
    print(f"Loaded existing results: {len(all_results['accounts'])} accounts already scanned")

done_handles = set(a["handle"] for a in all_results.get("accounts", []))

for i, handle in enumerate(HANDLES):
    if handle in done_handles:
        print(f"[{i+1}/{len(HANDLES)}] {handle} — SKIP (already scanned)")
        continue
    
    print(f"[{i+1}/{len(HANDLES)}] {handle}...", end=" ", flush=True)
    result = fetch_tweets(handle)
    
    if "error" in result:
        print(f"ERROR: {result['error']}")
        all_results.setdefault("errors", []).append({"handle": handle, "error": result["error"]})
    else:
        raw_tweets = result["tweets"]
        
        # Apply date filter
        recent = []
        for t in raw_tweets:
            ca = t.get("createdAt", "")
            dt = parse_date(ca)
            if dt and dt >= CUTOFF:
                autor = t.get("author", {})
                tweet_data = {
                    "id": t.get("id", ""),
                    "text": t.get("text", ""),
                    "createdAt": ca,
                    "date": dt.strftime("%Y-%m-%d") if dt else "",
                    "likes": t.get("likeCount", 0) or 0,
                    "retweets": t.get("retweetCount", 0) or 0,
                    "replies": t.get("replyCount", 0) or 0,
                    "views": t.get("viewCount", 0) or 0,
                    "bookmarks": t.get("bookmarkCount", 0) or 0,
                    "author_name": autor.get("userName", handle),
                    "author_followers": autor.get("followers", 0) or 0,
                }
                tweet_data["score"] = (
                    tweet_data["likes"] +
                    tweet_data["retweets"] * 2 +
                    tweet_data["replies"] +
                    tweet_data["views"] / 100
                )
                recent.append(tweet_data)
        
        qualifying = [t for t in recent if 
                      t["likes"] >= 50 or t["retweets"] >= 10 or t["views"] >= 1000]
        qualifying.sort(key=lambda x: x["score"], reverse=True)
        
        is_filtered = handle in IRRELEVANT
        account_data = {
            "handle": handle,
            "bio": "",
            "filtered": is_filtered,
            "total_fetched": len(raw_tweets),
            "recent_count": len(recent),
            "qualifying_count": len(qualifying),
            "tweets": qualifying,
        }
        all_results["accounts"].append(account_data)
        
        sample_dates = [t["date"] for t in recent[:3]] if recent else []
        filter_tag = " [FILTERED]" if is_filtered else ""
        print(f"{len(raw_tweets)} fetched, {len(recent)} recent{', dates=' + ','.join(sample_dates) if sample_dates else ''}, {len(qualifying)} qualifying{filter_tag}")
    
    with open(RESULTS_FILE + ".tmp", "w") as f:
        json.dump(all_results, f, indent=2)
    
    time.sleep(1)

with open(RESULTS_FILE, "w") as f:
    json.dump(all_results, f, indent=2)
os.remove(RESULTS_FILE + ".tmp") if os.path.exists(RESULTS_FILE + ".tmp") else None

total_q = sum(a["qualifying_count"] for a in all_results["accounts"])
print(f"\n=== Chunk 2 Complete: {len(HANDLES)} scanned, {total_q} total qualifying tweets ===")