from pathlib import Path
import sqlite3
import pandas as pd

PROJECT_ROOT = Path(__file__).resolve().parent.parent
DB_PATH = PROJECT_ROOT / "data" / "hiffi_catalog.db"

conn = sqlite3.connect(DB_PATH)
query = """
    SELECT 
        a.artist_id, a.artist_name, a.youtube_url, 
        t.subscribers_count, t.spotify_monthly_listeners
    FROM artists a
    JOIN telemetry_metrics t ON a.artist_id = t.artist_id
    WHERE a.image_verified_flag = 1
"""
df = pd.read_sql_query(query, conn)
conn.close()

# Flag remaining search queries
still_search = df[df["youtube_url"].astype(str).str.contains(r"results\?search_query", regex=True, na=False)]

# Flag suspicious subscriber counts (< 5,000 subs for artists with > 200,000 Spotify listeners)
mismatched_subs = df[(df["spotify_monthly_listeners"] > 200000) & (df["subscribers_count"] < 5000)]

print(f"Remaining YouTube search fallback URLs: {len(still_search)}")
print(f"Flagged subscriber mismatches (< 5k subs with high Spotify reach): {len(mismatched_subs)}")

if not mismatched_subs.empty:
    print("\nSuspicious channel mappings needing verified override:")
    print(mismatched_subs[["artist_name", "subscribers_count", "spotify_monthly_listeners", "youtube_url"]].to_string(index=False))