import nmap
import pandas as pd
import os
import sys
import requests
from bs4 import BeautifulSoup
import urllib3
from datetime import datetime

# Suppress SSL warnings for self-signed certificates
urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)

def probe_for_web_interface(host, port):
    """
    Zero-trust probe: Manually tries HTTPS and HTTP on EVERY open port.
    Captures webpage title and software version banners.
    """
    for protocol in ["https", "http"]:
        url = f"{protocol}://{host}:{port}"
        try:
            headers = {'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64)'}
            response = requests.get(url, timeout=3, verify=False, allow_redirects=True, headers=headers)
            
            # We accept any response < 500 (200, 302, 401, 403)
            if response.status_code < 500:
                soup = BeautifulSoup(response.text, 'html.parser')
                title = soup.title.string.strip() if soup.title else "Web Page (No Title)"
                # Grab the 'Server' header to identify the software stack
                server_banner = response.headers.get('Server', 'Hidden/Unknown')
                return True, url, title, server_banner
        except:
            continue 
    return False, None, None, None

def run_exhaustive_scan(target_range):
    nm = nmap.PortScanner()
    
    # PHASE 1: Rapid Discovery
    print(f"[*] PHASE 1: Discovering live hosts on {target_range}...")
    nm.scan(hosts=target_range, arguments='-sn')
    live_hosts = nm.all_hosts()
    
    if not live_hosts:
        print("[!] No devices found. Check your connection to the 192.168.12.x VLAN.")
        return []

    print(f"[+] Found {len(live_hosts)} active devices.")
    print("[*] PHASE 2: Deep Scanning top 10,000 ports (this will take a few minutes)...")
    
    # Using synchronous scan for Windows stability
    target_list = " ".join(live_hosts)
    # -sV: Service version detection for hostnames and initial banners
    nm_args = '-sV --open -T4 --top-ports 10000'
    nm.scan(hosts=target_list, arguments=nm_args)

    # PHASE 3: Zero-Trust Web Verification & Banner Grabbing
    scan_results = []
    print("\n[*] PHASE 3: Verifying web interfaces and grabbing software banners...")
    
    for host in nm.all_hosts():
        hostname = nm[host].hostname()
        print(f"\n[+] Analyzing host: {host} ({hostname if hostname else 'No DNS Name'})")
        
        for proto in nm[host].all_protocols():
            lport = nm[host][proto].keys()
            for port in lport:
                print(f"    - Probing port {port}...")
                is_web, verified_url, web_title, banner = probe_for_web_interface(host, port)
                
                if is_web:
                    scan_results.append({
                        'IP Address': host,
                        'Hostname': hostname,
                        'Port': port,
                        'Service Identity': web_title,
                        'Software Banner': banner,
                        'Verified Login URL': verified_url,
                        'Discovery Date': datetime.now().strftime("%Y-%m-%d %H:%M")
                    })
                    
    return scan_results

def save_to_excel(data, output_path):
    if not data:
        print("\n[!] No web services detected on the identified hosts.")
        return

    df = pd.DataFrame(data)
    # XlsxWriter is used to ensure URLs are formatted as clickable blue links
    writer = pd.ExcelWriter(output_path, engine='xlsxwriter')
    df.to_excel(writer, index=False, sheet_name='Homelab_Inventory')
    
    workbook = writer.book
    worksheet = writer.sheets['Homelab_Inventory']
    
    # Formatting for better readability
    for i, col in enumerate(df.columns):
        width = max(df[col].astype(str).map(len).max(), len(col)) + 2
        worksheet.set_column(i, i, width)
        
    writer.close()
    print(f"\n[SUCCESS] Discovery Finished! Your documentation is at: {output_path}")

if __name__ == "__main__":
    # Settings
    TARGET_VLAN = "192.168.12.0/24"
    # Storing in your requested directory
    OUTPUT_FILE = r"D:\hackerstuff\scripts\my_scripts\ai-code\Full_Homelab_Recovery.xlsx"

    os.makedirs(os.path.dirname(OUTPUT_FILE), exist_ok=True)
    results = run_exhaustive_scan(TARGET_VLAN)
    save_to_excel(results, OUTPUT_FILE)