Implement multiprocessing for brute-force password attempts

[bloodstiller]

Refactor bfkeepass.py to use multiprocessing for password attempts and improve argument parsing.

Summary

Refactors bfkeepass to improve bruteforce performance through parallelism, add real-time verbose progress reporting, and harden error handling.

---

Changes

Performance

• Replaced sequential password attempts with ProcessPoolExecutor for true multiprocessing — KeePass KDF (Argon2/AES) is CPU-bound so processes bypass the GIL where threads cannot
• Added -t / --threads argument to control worker count (default: 4)
• Wordlist is consumed in configurable batches (BATCH_SIZE = 200) to avoid loading large wordlists (e.g. rockyou.txt) fully into memory

Early Termination

• Added a multiprocessing.Manager shared Event (stop_flag) passed to all workers
• Workers check the flag at the start of each attempt and exit immediately once a password is found, preventing wasted CPU cycles across all processes

Verbose Output

• Moved all print calls to the main process — worker process stdout is unreliable across process boundaries
• Verbose mode now prints each attempted password with its attempt index in real time using flush=True
• Final summary line reports total passwords attempted regardless of outcome

Error Handling

• Replaced bare except: with explicit CredentialsError catch — the expected failure path for a wrong password
• Unexpected exceptions in workers are now surfaced as [WARN] messages to stderr rather than silently swallowed
• Wordlist open errors now distinguish FileNotFoundError from general OSError and exit with code 1

Code Quality

• Removed type=ascii from argparse arguments and the subsequent .replace("'", "") workaround
• Extracted dump_entries() and iter_batches() as standalone functions
• Workers serialize PyKeePass entries to plain dicts before returning — PyKeePass objects are not picklable across process boundaries
• Added type hints throughout

Verified On Test KDBX