keyhunter-parser/wordpress_eternal.py

"""ETERNAL WordPress.org plugin firehose: plugins API + zip scan.

Anonymous. Browse rotation: updated -> new -> popular (per cycle), 5 pages
of 100. Each new plugin: download zip (cap 8MB), scan text files via
key_patterns.scan_text. seen=slug. mark_seen ONLY on success.
"""
import os
KF_BASE = os.path.dirname(os.path.abspath(__file__))
os.makedirs(os.path.join(KF_BASE, "data"), exist_ok=True)
import hashlib
import io
import sqlite3
import sys
import threading
import time
import zipfile

import requests

DB = os.environ.get("KF_DB") or os.path.join(KF_BASE, "data/keys.db")
LOG = os.path.join(KF_BASE, "data/wp_eternal.log")
SEEN_FILE = os.path.join(KF_BASE, "data/wp_seen.txt")
N_THREADS = 6
PAGES_PER_CYCLE = 5
ZIP_CAP = 8 * 1024 * 1024
FILE_CAP = 500_000
SLEEP = 90
BROWSES = ("updated", "new", "popular")

from key_patterns import scan_text

GARBAGE = ("xxxx", "your", "YOUR", "EXAMPLE", "example", "n0tr3al",
           "placeholder", "CHANGEME", "changeme", "dummy", "DUMMY",
           "abcdefghijklmnopqrst", "1234567890abcdef")

SCAN_EXT = (".php", ".js", ".json", ".txt", ".env", ".xml", ".yaml", ".yml",
            ".ini", ".cfg", ".po", ".csv")

UA = {"User-Agent": "keyfarm-wp/1.0"}


def log(msg):
    line = f"[{time.strftime('%Y-%m-%d %H:%M:%S')}] {msg}\n"
    with open(LOG, "a", encoding="utf-8") as f:
        f.write(line)


conn = sqlite3.connect(DB, timeout=120, check_same_thread=False,
                       isolation_level=None)
lock = threading.Lock()
total = 0


def db_add(key, prov, repo):
    global total
    with lock:
        conn.execute(
            "INSERT INTO keys (hash, val, prov, status, found, repo) "
            "VALUES (?,?,?,?,?,?) "
            "ON CONFLICT(hash) DO UPDATE SET status='NEW', "
            "found=excluded.found, repo=excluded.repo "
            "WHERE keys.status IN ('DEAD','ERR')",
            (hashlib.sha1(key.encode()).hexdigest()[:16],
             key, prov, "NEW", "wp", repo))
        total += 1


def load_seen():
    s = set()
    try:
        with open(SEEN_FILE, encoding="utf-8") as f:
            for line in f:
                line = line.strip()
                if line:
                    s.add(line)
    except FileNotFoundError:
        pass
    return s


seen = load_seen()


def mark_seen(p):
    seen.add(p)
    with open(SEEN_FILE, "a", encoding="utf-8") as f:
        f.write(p + "\n")


def work(slug, dl):
    """Fetch plugin zip, scan. -1 on final fail."""
    n = 0
    for attempt in (1, 2, 3):
        try:
            r = requests.get(dl, headers=UA, timeout=40, stream=True)
            if r.status_code == 404:
                return 0
            if r.status_code != 200:
                raise RuntimeError(f"http {r.status_code}")
            cl = int(r.headers.get("Content-Length") or 0)
            if cl > ZIP_CAP:
                return 0
            buf = r.content
            if len(buf) > ZIP_CAP:
                return 0
            zf = zipfile.ZipFile(io.BytesIO(buf))
            for m in zf.infolist():
                if m.is_dir() or m.file_size > FILE_CAP:
                    continue
                fn = m.filename.lower()
                if not any(fn.endswith(e) for e in SCAN_EXT):
                    continue
                text = zf.read(m).decode("utf-8", errors="ignore")
                for prov, key in scan_text(text):
                    if isinstance(key, tuple):
                        key = key[0]
                    if any(g in key for g in GARBAGE) or len(key) < 30:
                        continue
                    db_add(key, prov, f"wordpress:{slug}")
                    n += 1
            return n
        except Exception:
            if attempt == 3:
                log(f"  wp-err {slug}")
                return -1
            time.sleep(8)
    return -1


def cycle():
    global total
    total = 0
    browse = BROWSES[int(time.time() / SLEEP) % len(BROWSES)]
    items = []
    for page in range(1, PAGES_PER_CYCLE + 1):
        params = {"action": "query_plugins", "request[per_page]": "100",
                  "request[page]": str(page), "request[browse]": browse}
        for attempt in (1, 2, 3):
            try:
                r = requests.get("https://api.wordpress.org/plugins/info/1.2/",
                                 params=params, headers=UA, timeout=40)
                if r.status_code != 200:
                    raise RuntimeError(f"http {r.status_code}")
                d = r.json()
                break
            except Exception as e:
                if attempt == 3:
                    log(f"list-err browse={browse} page={page} "
                        f"{type(e).__name__}")
                    return
                time.sleep(8)
        for p in d.get("plugins") or []:
            slug = p.get("slug")
            dl = p.get("download_link")
            if slug and dl and slug not in seen:
                items.append((slug, dl))

    ok = fail = 0
    from concurrent.futures import ThreadPoolExecutor, as_completed
    with ThreadPoolExecutor(max_workers=N_THREADS) as ex:
        futs = {ex.submit(work, s, dl): s for s, dl in items}
        for fut in as_completed(futs):
            s = futs[fut]
            try:
                res = fut.result()
            except Exception:
                res = -1
            if res == -1:
                fail += 1          # NOT marked seen: retried next cycle
            else:
                ok += 1
                mark_seen(s)

    log(f"cycle: browse={browse} items={len(items)} ok={ok} "
        f"fail={fail} +{total} keys")


def main():
    log("wp_eternal started")
    if "--once" in sys.argv:
        cycle()
        return
    while True:
        try:
            cycle()
        except Exception as e:
            log(f"cycle-err {type(e).__name__}: {e}")
        time.sleep(SLEEP)


if __name__ == "__main__":
    main()