KeyHunter parser v10 + 18 scanners + tests (synthetic fixtures only, no keys/db/tokens)

crates_eternal.py

@@ -0,0 +1,152 @@
+"""ETERNAL crates.io hunt: registry search + .crate tarball unpack + grep."""
+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 json
+import re
+import sqlite3
+import tarfile
+import time
+import urllib.parse
+import urllib.error
+import urllib.request
+
+DB = os.environ.get("KF_DB") or os.path.join(KF_BASE, "data/keys.db")
+LOG = os.path.join(KF_BASE, "crates_eternal.log")
+UA = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"}
+UA["Accept"] = "application/json"
+
+QUERIES = [
+    "openai", "anthropic", "claude", "dashscope", "openrouter", "deepseek",
+    "groq", "zhipu", "glm", "moonshot", "kimi", "together", "fireworks",
+    "replicate", "nvidia nim", "siliconflow", "telegram bot", "llm wrapper",
+    "ai sdk", "chatbot", "gpt wrapper", "agent framework", "voice ai",
+    "tts", "embedding", "api key", "secret", "config env",
+]
+
+SCAN_EXT = (".rs", ".toml", ".json", ".md", ".env", ".yml", ".yaml", ".txt")
+
+from key_patterns import PATTERNS
+
+GARBAGE = ("xxxx", "your", "EXAMPLE", "example", "n0tr3al", "INSERT", "FAKE",
+           "fake", "Fake", "test", "Test", "TEST", "demo", "Demo", "aaaa",
+           "aBcDeF", "abcdef", "1234567890", "abcdefghij", "placehol",
+           "changeme", "deadbeef", "fffffff", "0000000")
+
+
+def log(msg):
+    line = f"[{time.strftime('%Y-%m-%d %H:%M:%S')}] {msg}\n"
+    print(line, flush=True)
+    with open(LOG, "a", encoding="utf-8", errors="replace") as f:
+        f.write(line)
+
+
+def db_add(cur, seen, key, prov, repo):
+    if key in seen or any(g in key for g in GARBAGE) or len(key) < 28:
+        return 0
+    seen.add(key)
+    cur.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",
+         "crates", repo))
+    return 1
+
+
+def scan_tarball(url, cur, seen, pkg):
+    n = 0
+    try:
+        req = urllib.request.Request(url, headers=UA)
+        with urllib.request.urlopen(req, timeout=40) as r:
+            blob = r.read()
+    except Exception as e:
+        log(f"  crate {pkg} err: {str(e)[:50]}")
+        return 0
+    if len(blob) > 5_000_000:
+        return 0
+    try:
+        tf = tarfile.open(fileobj=io.BytesIO(blob), mode="r:gz")
+        members = tf.getmembers()
+    except Exception as e:
+        log(f"  untar {pkg} err: {str(e)[:50]}")
+        return 0
+    for m in members:
+        if m.size > 600_000:
+            continue
+        name = m.name.lower()
+        if not name.endswith(SCAN_EXT) and "/src/" not in name:
+            if not name.endswith(SCAN_EXT):
+                continue
+        try:
+            fdata = tf.extractfile(m)
+            if fdata is None:
+                continue
+            data = fdata.read().decode(errors="replace")
+        except Exception:
+            continue
+        for prov, rx in PATTERNS:
+            for key in rx.findall(data):
+                if isinstance(key, tuple):
+                    key = key[-1]
+                hit = db_add(cur, seen, key, prov, f"crates:{pkg}")
+                if hit:
+                    n += 1
+                    log(f"  + {prov} {key[:24]}... crates:{pkg}/{m.name}")
+    return n
+
+
+def cycle(db, qi):
+    cur = db.cursor()
+    cur.execute("SELECT val FROM keys")
+    seen = {r[0] for r in cur.fetchall()}
+    total = 0
+    q = QUERIES[qi % len(QUERIES)]
+    qi += 1
+    try:
+        req = urllib.request.Request(
+            f"https://crates.io/api/v1/crates?q={urllib.parse.quote(q)}&per_page=50",
+            headers=UA)
+        with urllib.request.urlopen(req, timeout=30) as r:
+            d = json.loads(r.read())
+        crates = d.get("crates", [])
+        log(f"query {q!r}: {len(crates)} crates")
+        for item in crates[:25]:
+            name = item.get("name", "")
+            ver = item.get("max_version", "") or item.get("newest_version", "")
+            if not name or not ver:
+                continue
+            tb = f"https://static.crates.io/crates/{name}/{name}-{ver}.crate"
+            try:
+                total += scan_tarball(tb, cur, seen, name)
+            except Exception as e:
+                log(f"  crate {name} err: {str(e)[:40]}")
+            time.sleep(1)
+    except urllib.error.HTTPError as e:
+        log(f"crates.io HTTP {e.code} for {q!r}; sleeping 5min")
+        db.commit()
+        time.sleep(300)
+    except Exception as e:
+        log(f"cycle err: {str(e)[:80]}")
+    db.commit()
+    return total, qi
+
+
+def main():
+    log(f"=== crates eternal START pid={__import__('os').getpid()} ===")
+    db = sqlite3.connect(DB, timeout=120, check_same_thread=False, isolation_level=None)
+    qi = 0
+    while True:
+        try:
+            n, qi = cycle(db, qi)
+            log(f"CYCLE done: +{n}; resting 3min")
+        except Exception as e:
+            log(f"cycle error (continuing): {e}")
+        time.sleep(180)
+
+
+if __name__ == "__main__":
+    main()