Fix/dnsdumpster csrf robust and subdomain scan

[eMi-]

This PR fixes two issues:

1. DNSDumpster CSRF parsing & blocking

   • Replaces brittle token extraction with tolerant parsing (input + regex fallback).
   • Adds realistic Session headers (UA + Referer).
   • Returns a structured empty result on failure instead of [] to avoid type errors.
   • Emits a clear message ("CSRF token not found…") when blocked (captcha/Tor/VPN).

2. Subdomain scan progress ZeroDivisionError

   • When the wordlist has < 100 entries, the progress step calculation produced 0 → c % 0 crash.
   • Use step = max(1, int(numOfLines/100.0)) to prevent modulo-by-zero.
   • Also reads the same file for count + scan, adds small timeouts on HTTP checks.

Tested on Parrot/Kali (Python3), with/without Tor.
No behavior change on successful paths; only more robust error handling.