Extractor.py

import cv2
import numpy as np
import base64
import PyPDF2
import os
from cryptography.hazmat.primitives import hashes
from cryptography.hazmat.primitives.kdf.pbkdf2 import PBKDF2HMAC
from cryptography.fernet import Fernet

# --- CORE TOOLS (MUST MATCH SENDER) ---
def get_crypto_tools(password: str):
    salt = b'\x00' * 16 
    kdf = PBKDF2HMAC(algorithm=hashes.SHA256(), length=32, salt=salt, iterations=100000)
    key = base64.urlsafe_b64encode(kdf.derive(password.encode()))
    return Fernet(key)

# --- RECOVERY ENGINE ---

def recover_from_image(stego_image, output_file, password, is_text=False):
    img = cv2.imread(stego_image)
    if img is None: return
    flat_bits = [str(p & 1) for p in img.flatten()]
    full_bin = "".join(flat_bits)
    all_bytes = bytearray()
    for i in range(0, len(full_bin), 8):
        byte = int(full_bin[i:i+8], 2)
        all_bytes.append(byte)
        if all_bytes.endswith(b"STOP_HERE"): break
    cipher = get_crypto_tools(password)
    decrypted = cipher.decrypt(bytes(all_bytes[:-9]))
    if is_text: 
        print(f"🔓 TEXT FROM IMAGE: {decrypted.decode()}")
    else:
        with open(output_file, 'wb') as f: f.write(decrypted)
        print(f"🔓 FILE FROM IMAGE RECOVERED: {output_file}")

def recover_from_pdf(stego_pdf, output_file, password, is_text=False):
    reader = PyPDF2.PdfReader(stego_pdf)
    encrypted_blob = reader.metadata.get('/Subject')
    cipher = get_crypto_tools(password)
    decrypted = cipher.decrypt(encrypted_blob.encode())
    if is_text: 
        print(f"🔓 TEXT FROM PDF: {decrypted.decode()}")
    else:
        with open(output_file, 'wb') as f: f.write(decrypted)
        print(f"🔓 FILE FROM PDF RECOVERED: {output_file}")

def recover_from_file(stego_file, output_file, password, is_text=False):
    with open(stego_file, 'rb') as f: data = f.read()
    if b"FILE_SECRET_START" not in data: return
    secret_part = data.split(b"FILE_SECRET_START")[1]
    cipher = get_crypto_tools(password)
    decrypted = cipher.decrypt(secret_part)
    if is_text: 
        print(f"🔓 TEXT FROM FILE: {decrypted.decode()}")
    else:
        with open(output_file, 'wb') as f: f.write(decrypted)
        print(f"🔓 FILE FROM FILE RECOVERED: {output_file}")

# --- PRODUCTION RECOVERY (THE DEMO) ---
PASS = "ismagi_secure_2026"

# --- PRODUCTION RECOVERY (THE DEMO) ---
PASS = "ismagi_secure_2026"
# This is the folder where your hidden files are sitting
HIDDEN_DIR = "the hidden stuff"

# --- FULL PRODUCTION RECOVERY ---
PASS = "ismagi_secure_2026"
HIDDEN_DIR = "the hidden stuff"

# --- FULL PRODUCTION RECOVERY (A through I) ---
PASS = "ismagi_secure_2026"
HIDDEN_DIR = "the hidden stuff"

print("--- STARTING FULL A-I EXTRACTION ---")

# A & H: Photo Hidden in Photo
# recover_from_image(os.path.join(HIDDEN_DIR, "image_in_image.png"), "RECOVERED_A_PHOTO_FROM_IMAGE.png", PASS)

# B. Typed Text inside Photo
recover_from_image(os.path.join(HIDDEN_DIR, "text_in_image.png"), None, PASS, is_text=True)

# # C. File (message.txt) inside PDF
# recover_from_pdf(os.path.join(HIDDEN_DIR, "message_in_pdf.pdf"), "RECOVERED_C_MESSAGE_FROM_PDF.txt", PASS)

# # D. Typed Text inside PDF
# recover_from_pdf(os.path.join(HIDDEN_DIR, "text_in_pdf.pdf"), None, PASS, is_text=True)

# # E. PDF inside Text File
# recover_from_file(os.path.join(HIDDEN_DIR, "pdf_hidden_in_text.txt"), "RECOVERED_E_PDF_FROM_TEXT.pdf", PASS)

# # F. Typed Text inside Text File
# recover_from_file(os.path.join(HIDDEN_DIR, "hidden_in_text.txt"), None, PASS, is_text=True)

# # G. Photo inside Text File
# recover_from_file(os.path.join(HIDDEN_DIR, "photo_hidden_in_text.txt"), "RECOVERED_G_PHOTO_FROM_TEXT.png", PASS)

# # H. PDF inside Photo (Duplicate logic of A, but with different filename)
# recover_from_image(os.path.join(HIDDEN_DIR, "pdf_in_photo.png"), "RECOVERED_H_PDF_FROM_PHOTO.pdf", PASS)

# # I. PDF inside PDF
# recover_from_pdf(os.path.join(HIDDEN_DIR, "pdf_in_pdf.pdf"), "RECOVERED_I_PDF_FROM_PDF.pdf", PASS)

print("--- ALL 9 SCENARIOS RECOVERED ---")