hiffi-artist-intelligence/scripts/resolve_youtube_verified_oac.py

"""
scripts/resolve_youtube_verified_oac.py
Resolves the remaining unresolved YouTube channels using a curated canonical
handle dictionary and enforces a strict database safety floor.
"""

import json
import re
import sys
import time
from pathlib import Path
import requests

ROOT_DIR = Path(__file__).resolve().parent.parent
if str(ROOT_DIR) not in sys.path:
    sys.path.insert(0, str(ROOT_DIR))

from core.database import get_connection

HEADERS = {
    "User-Agent": (
        "Mozilla/5.0 (Windows NT 10.0; Win64; x64) "
        "AppleWebKit/537.36 (KHTML, like Gecko) "
        "Chrome/124.0.0.0 Safari/537.36"
    ),
    "Accept-Language": "en-US,en;q=0.9",
}

# Curated handles for artists where guessing @artistname fails
CANONICAL_HANDLES = {
    "Pusha T": "kingpush",
    "Flatbush Zombies": "FlatbushZombiesTV",
    "Larry June": "LarryJuneTFM",
    "Curren$y": "CurrensyJetLife",
    "Jeezy": "JeezyOfficial",
    "Smoke DZA": "SmokeDZA",
    "Jay-Z": "JayZVEVO",
    "Juvenile": "JuvenileVEVO",
    "Bun B": "BunBVEVO",
    "Young Buck": "YoungBuckVEVO",
    "Twista": "TwistaVEVO",
    "Cam'ron": "CamronVEVO",
    "Styles P": "StylesPVEVO",
    "Chingy": "ChingyVEVO",
    "Kool G Rap": "KoolGRapOfficial",
    "KRS-One": "KRSOneOfficial",
    "The Roots": "TheRootsVEVO",
    "Inspectah Deck": "InspectahDeckVEVO",
    "Juicy J": "JuicyJVEVO",
    "Schoolboy Q": "SchoolboyQVEVO",
    "Ab-Soul": "TopDawgEnt",
    "Quavo": "QuavoOfficial",
    "Slim Jxmmi": "RaeSremmurd",
    "Rich Homie Quan": "RichHomieQuanVEVO",
    "Key Glock": "KeyGlockVEVO",
    "Kevin Gates": "kevingatesTV",
    "Boosie Badazz": "BoosieBadazzVEVO",
    "King Von": "KingVon",
    "Trippie Redd": "TrippieRedd",
    "Ski Mask the Slump God": "SkiMaskTheSlumpGodVEVO",
    "Cordae": "cordae",
    "Mick Jenkins": "MickJenkinsVEVO",
    "Benny The Butcher": "BennyTheButcherVEVO",
    "Boldy James": "BoldyJamesVEVO",
    "Roc Marciano": "RocMarciano",
    "Maxo Kream": "MaxoKreamVEVO",
    "Big K.R.I.T.": "BigKRITVEVO",
    "Jay Worthy": "JayWorthyVEVO",
    "Veeze": "VeezeOfficial",
    "YN Jay": "YNJay",
    "Fredo Santana": "FredoSantana300",
    "Jay Critch": "JayCritchVEVO",
    "Saweetie": "OfficialSaweetie",
    "Yung Miami": "CityGirlsVEVO",
    "Tierra Whack": "TierraWhack",
    "Dizzee Rascal": "DizzeeRascalVEVO",
    "Digga D": "DiggaD",
    "K-Trap": "KTrapOfficial",
    "M Huncho": "MHunchoOfficial",
    "Nines": "NinesVEVO",
    "Potter Payper": "PotterPayperOfficial",
    "Rimzee": "RimzeeOfficial",
    "MoStack": "MoStackVEVO",
    "Stefflon Don": "StefflonDonVEVO",
    "Summrs": "SummrsArchive",
    "BabyTron": "BabyTronOfficial",
    "Luh Tyler": "LuhTylerVEVO",
    "Rob49": "Rob49VEVO",
    "Hit-Boy": "HitBoyVEVO",
    "Lil Skies": "LilSkies",
    "Smokepurpp": "SmokepurppVEVO",
    "Lil Pump": "LilPump",
    "Chris Travis": "ChrisTravisVEVO",
    "Yung Lean": "YungLeanVEVO",
    "Bladee": "DrainGang",
    "Ecco2k": "DrainGang",
    "Thaiboy Digital": "DrainGang",
    "BBNO$": "bbnomula",
    "Jhené Aiko": "JheneAiko",
    "Tinashe": "TinasheOfficial",
    "Brent Faiyaz": "BrentFaiyaz",
    "Lucky Daye": "LuckyDayeVEVO",
    "Meechy Darko": "MeechyDarkoVEVO",
    "NoCap": "NoCap16_29",
    "Rylo Rodriguez": "RyloRodriguezVEVO",
    "Lil Baby": "LilBabyVEVO",
    "EARTHGANG": "EarthgangVEVO",
    "Mobb Deep": "MobbDeepVEVO",
    "David Banner": "DavidBannerVEVO",
    "Lil Keke": "LilKekeVEVO",
    "Reason": "TopDawgEnt",
    "Foolio": "JulioFoolio",
    "Chika": "ChikaOfficial",
    "Fredo": "FredoOfficial",
    "AG Club": "AGClubOfficial",
    "Pouya": "PouyaKevin",
    "BONES": "TeamSESH",
    "Erick the Architect": "ErickTheArchitect",
    "Shy Glizzy": "ShyGlizzyVEVO",
    "Autumn!": "Twinuzis",
    "SpaceGhostPurrp": "SpaceGhostPurrp",
    "Mos Def": "MosDefVEVO",
    "SiR": "TopDawgEnt",
    "Homixide Gang": "HomixideGang",
    "BabySmoove": "BabySmooveVEVO",
}


def parse_metric(raw_text):
    if not raw_text:
        return 0
    t = str(raw_text).upper().replace(",", "").strip()
    match = re.search(r"([\d\.]+)\s*([KMB])?", t)
    if not match:
        return 0
    val = float(match.group(1))
    suffix = match.group(2)
    if suffix == "K":
        val *= 1e3
    elif suffix == "M":
        val *= 1e6
    elif suffix == "B":
        val *= 1e9
    return int(val)


def extract_subscribers_from_html(html_text):
    match = re.search(r"var ytInitialData\s*=\s*({.+?});</script>", html_text)
    if not match:
        match = re.search(r"window\[\"ytInitialData\"\]\s*=\s*({.+?});", html_text)

    if match:
        try:
            data = json.loads(match.group(1))
            header = data.get("header", {})
            c4 = header.get("c4TabbedHeaderRenderer", {})
            sub_text = c4.get("subscriberCountText", {}).get("simpleText", "")
            if not sub_text:
                runs = c4.get("subscriberCountText", {}).get("runs", [])
                sub_text = "".join(r.get("text", "") for r in runs)
            if sub_text:
                return parse_metric(sub_text)

            vm = (
                header.get("pageHeaderRenderer", {})
                .get("content", {})
                .get("pageHeaderViewModel", {})
                .get("metadata", {})
                .get("contentMetadataViewModel", {})
                .get("metadataRows", [])
            )
            for row in vm:
                for part in row.get("metadataParts", []):
                    text = part.get("text", {}).get("content", "")
                    if "subscribers" in text.lower():
                        return parse_metric(text)
        except Exception:
            pass

    meta_match = re.search(r'([\d\.]+[KMB]?)\s+subscribers', html_text, re.IGNORECASE)
    if meta_match:
        return parse_metric(meta_match.group(1))

    return 0


def fetch_handle_subs(handle):
    url = f"https://www.youtube.com/@{handle}"
    try:
        r = requests.get(url, headers=HEADERS, timeout=10)
        if r.status_code == 200:
            return extract_subscribers_from_html(r.text)
    except Exception:
        pass
    return 0


def main():
    print("=" * 65)
    print("HIFFI INTELLIGENCE: CANONICAL YOUTUBE RESOLUTION & SAFETY AUDIT")
    print("=" * 65)

    conn = get_connection()
    cur = conn.cursor()

    updated = 0
    preserved = 0

    for artist, handle in CANONICAL_HANDLES.items():
        # Get existing count
        cur.execute("""
            SELECT t.subscribers_count 
            FROM telemetry_metrics t
            JOIN artists a ON t.artist_id = a.artist_id
            WHERE LOWER(a.artist_name) = LOWER(?)
        """, (artist,))
        row = cur.fetchone()
        existing_subs = row[0] if row and row[0] else 0

        scraped_subs = fetch_handle_subs(handle)

        # Enforce safety floor
        if scraped_subs > 0:
            if scraped_subs >= existing_subs or existing_subs in [0, 500000]:
                cur.execute("""
                    UPDATE telemetry_metrics
                    SET subscribers_count = ?
                    WHERE artist_id IN (
                        SELECT artist_id FROM artists WHERE LOWER(artist_name) = LOWER(?)
                    )
                """, (scraped_subs, artist))
                print(f"[UPDATED]   {artist:<22} -> {scraped_subs:,} subs (via @{handle})")
                updated += 1
            else:
                print(f"[PRESERVED] {artist:<22} -> Kept {existing_subs:,} subs (Scraped @{handle} was {scraped_subs:,})")
                preserved += 1
        else:
            print(f"[FAILED]    {artist:<22} -> Handle @{handle} returned 0 / 404")

        time.sleep(0.4)

    conn.commit()
    conn.close()

    print("=" * 65)
    print(f"Complete: Updated {updated} channels | Preserved {preserved} verified baselines.")


if __name__ == "__main__":
    main()