My experimental hash algorithm (for learning purposes)

WHash.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "5d0af8bb-106f-4081-b5c3-51ab98cb4c45",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The message is: O cavalo morto eh um animal sem vida.\n",
      "The message bytes are: b'bRRQXlBGQFsaDBRQXlBGQFsaYxRQXlBGQFsaIBRQXlBGQFsaIhRQXlBGQFsaIgNQXlBGQFsaIhRQXlBGQFsaIhlQXlBGQFsaIhpQXlBGQFsaIhofXlBGQFsaIhpSXlBGQFsaIhpQXlBGQFsaIhpNXlBGQFsaIhpNWFBGQFsaIhpNQ1BGQFsaIhpNDFBGQFsaIhpNSVBGQFsaIhpNSV1GQFsaIhpNSRVGQFsaIhpNSUBGQFsaIhpNSVhGQFsaIhpNSVgLQFsaIhpNSVhKQFsaIhpNSVhFQFsaIhpNSVhCQFsaIhpNSVhCRFsaIhpNSVhCSFsaIhpNSVhCRVsaIhpNSQ=='\n"
     ]
    }
   ],
   "source": [
    "import base64\n",
    "\n",
    "\n",
    "def get_whash(message_bytes: list) -> list:\n",
    "    MSG_SIZE = len(message_bytes)\n",
    "    pairs = []\n",
    "    for i in range(0, MSG_SIZE, 2):\n",
    "        b1 = message_bytes[i]\n",
    "        b2 = message_bytes[i + 1] if i < MSG_SIZE - 1 else 0\n",
    "        pairs.append(b2 ^ b1)\n",
    "    return pairs\n",
    "\n",
    "\n",
    "def whash(message: str):\n",
    "    HASH_SIZE=256\n",
    "    counter = 1\n",
    "    initial_message = message\n",
    "    while len(message) < 4 * HASH_SIZE:\n",
    "        message = message + initial_message[counter:] + initial_message[:counter]\n",
    "        counter+=1\n",
    "    message = message[:HASH_SIZE * 4]\n",
    "    message_bytes = list(message.encode('UTF-8'))\n",
    "    while len(message_bytes) > HASH_SIZE:\n",
    "        message_bytes = get_whash(message_bytes)\n",
    "    return message_bytes\n",
    "\n",
    "\n",
    "message = 'O cavalo morto eh um animal sem vida.'\n",
    "message_hash = whash(message)\n",
    "\n",
    "\n",
    "print('The message is:', message)\n",
    "print('The message bytes are:', base64.b64encode(bytes(message_hash)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "b06443fc-ac72-47aa-8917-d17828803a6f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "848516"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import lipsum\n",
    "text = lipsum.generate_paragraphs(1000)\n",
    "len(text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "be9d3dde-ff63-46dc-b632-745eb3b2f3c3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The message is: Restat locus huic disputationi vel maxime necessarius de amicitia, quam, si voluptas summum sit bonu\n",
      "The message bytes are: b'MFkKVE4CE1FEHU4GD1JFFxxVGgAWS1VOFgAbEUgMZkgeD0V8EVcFXE4RBUoPCFNDUAwSQhZLEwBDT0BvEF8XBAZbCxEEXxYISA5EXBxZWFJUVhxLUEROXBZvABoTD0VBEU5MUAJKEEgOHBFKDVhUTglGE1BNbxIZUAIAVV5QCEoKVg4aC1F4D0cUQh4RNRsAR3gPflpDQRFVHUIVHFkFTQ4AQ0gRChxJFhZOClwIShcXD3FYQlhmDw0YQVpTHAxCA0QHT0gAQnoMXRpGTAdFBAZIQA5cDhQWDVtXVRBORUBFXxMQU2YYQkZEXEFYCBseCFIUTQYOCExaHFYLG0YORg=='\n"
     ]
    }
   ],
   "source": [
    "message = text\n",
    "message_hash = whash(message)\n",
    "\n",
    "\n",
    "print('The message is:', message[:100])\n",
    "print('The message bytes are:', base64.b64encode(bytes(message_hash)))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.15"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}