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

packagist_eternal.py

@@ -0,0 +1,161 @@
+"""ETERNAL Packagist harvest: search.json + p2 metadata dist.zip scan.
+Lane: packagist. Polite sleep, 25 packages per cycle."""
+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 sys
+import time
+import urllib.parse
+import urllib.request
+import zipfile
+
+DB = os.environ.get("KF_DB") or os.path.join(KF_BASE, "data/keys.db")
+LOG = os.path.join(KF_BASE, "packagist_eternal.log")
+
+QUERIES = ["openai", "anthropic", "claude", "llm", "chatgpt", "gemini",
+           "mistral", "groq", "deepseek", "openrouter", "api",
+           "ai", "gpt", "sdk", "client", "secrets"]
+
+SCAN_EXT = (".php", ".env", ".json", ".yml", ".yaml", ".md", ".xml",
+            ".dist", ".txt", ".js", ".ts", ".py")
+MAX_FILE = 2 * 1024 * 1024
+UA = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64)"}
+
+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")
+
+seen = set()
+
+
+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, key, prov, repo):
+    if isinstance(key, tuple):
+        key = key[0]
+    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",
+         "packagist", repo))
+    return 1
+
+
+def fetch(url, binary=False):
+    req = urllib.request.Request(url, headers=UA)
+    with urllib.request.urlopen(req, timeout=60) as r:
+        blob = r.read()
+    return blob if binary else blob.decode("utf-8", "replace")
+
+
+def scan_zip(blob, cur, pkg):
+    n = 0
+    try:
+        zf = zipfile.ZipFile(io.BytesIO(blob))
+    except Exception:
+        return 0
+    for info in zf.infolist():
+        name = info.filename.lower()
+        if info.file_size > MAX_FILE:
+            continue
+        if not (name.endswith(SCAN_EXT) or ".env" in name):
+            continue
+        try:
+            text = zf.read(info).decode("utf-8", "replace")
+        except Exception:
+            continue
+        for prov, rx in PATTERNS:
+            for m in rx.findall(text):
+                n += db_add(cur, m, prov, f"{pkg}:{info.filename}")
+    return n
+
+
+def scan_package(pkg, cur):
+    try:
+        meta = json.loads(fetch(
+            "https://repo.packagist.org/p2/" + pkg + ".json"))
+    except Exception as e:
+        log(f"  meta {pkg} err: {str(e)[:60]}")
+        return 0
+    versions = meta.get("packages", {}).get(pkg, [])
+    if not versions:
+        return 0
+    dist = versions[0].get("dist", {})
+    url = dist.get("url")
+    if not url or dist.get("type") != "zip":
+        return 0
+    try:
+        blob = fetch(url, binary=True)
+    except Exception as e:
+        log(f"  zip {pkg} err: {str(e)[:60]}")
+        return 0
+    return scan_zip(blob, cur, pkg)
+
+
+def cycle(db):
+    cur = db.cursor()
+    total = 0
+    for q in QUERIES:
+        n = 0
+        url = ("https://packagist.org/search.json?q=" +
+               urllib.parse.quote(q))
+        try:
+            data = json.loads(fetch(url))
+        except Exception as e:
+            log(f"  search '{q}' err: {str(e)[:60]}")
+            continue
+        for p in data.get("results", [])[:25]:
+            n += scan_package(p.get("name", ""), cur)
+            db.commit()
+            time.sleep(1.0)
+        total += n
+        log(f"  '{q}': +{n} new")
+    return total
+
+
+def main():
+    log(f"=== packagist eternal START pid={__import__('os').getpid()} "
+        f"queries={len(QUERIES)}")
+    try:
+        con = sqlite3.connect(DB, timeout=120, check_same_thread=False, isolation_level=None)
+        cur = con.cursor()
+        cur.execute("SELECT val FROM keys WHERE found='packagist'")
+        for r in cur.fetchall():
+            seen.add(r[0])
+        con.close()
+    except Exception:
+        pass
+    db = sqlite3.connect(DB, timeout=120, check_same_thread=False, isolation_level=None)
+    while True:
+        try:
+            n = cycle(db)
+            log(f"CYCLE done +{n}")
+        except Exception as e:
+            log(f"cycle err: {repr(e)[:100]}")
+            time.sleep(30)
+        time.sleep(180)
+
+
+if __name__ == "__main__":
+    try:
+        main()
+    except KeyboardInterrupt:
+        sys.exit(0)