{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "90a312dd",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "7140f2ef",
   "metadata": {},
   "source": [
    "## Pseudonymization and Depseudonymization\n",
    "\n",
    "1. generate keys"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2e19922b",
   "metadata": {
    "vscode": {
     "languageId": "shellscript"
    }
   },
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "logprep pseudo generate -f ./analyst 1024\n",
    "logprep pseudo generate -f ./depseudo 2048\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "03df49f3",
   "metadata": {},
   "source": [
    "2. pseudonymize a string"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3989821b",
   "metadata": {
    "vscode": {
     "languageId": "shellscript"
    }
   },
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "logprep pseudo pseudonymize ./analyst.crt ./depseudo.crt mystring"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4f30f225",
   "metadata": {},
   "source": [
    "3. depseudonymize the string"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "48c96106",
   "metadata": {
    "vscode": {
     "languageId": "shellscript"
    }
   },
   "outputs": [],
   "source": [
    "%%bash\n",
    "\n",
    "logprep pseudo depseudonymize ./analyst.key ./depseudo.key  < cyphertext from above >"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "76df9e2e",
   "metadata": {},
   "source": [
    "## Pseudonymizer Processor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "86fee246",
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import uuid\n",
    "from logprep.factory import Factory\n",
    "from logprep.util.time import TimeParser\n",
    "from logprep.ng.event.log_event import LogEvent\n",
    "from logprep.ng.event.event_state import EventStateType\n",
    "import logging\n",
    "import sys\n",
    "\n",
    "# Configure logging\n",
    "logging.basicConfig(level=logging.DEBUG, stream=sys.stdout)\n",
    "\n",
    "document = {\n",
    "    \"id\": f\"{uuid.uuid4()}\",\n",
    "    \"@timestamp\": str(TimeParser.now()),\n",
    "    \"user\": {\n",
    "        \"name\": \"Hubert K. Kabal\",\n",
    "        \"email\": \"kabal@example.com\",\n",
    "        \"id\": 12345,\n",
    "    },\n",
    "}\n",
    "\n",
    "event = LogEvent(document, original=b\"\", state=EventStateType.RECEIVED)\n",
    "\n",
    "print(f\"Event before processing: {json.dumps(event.data, indent=2)}\")\n",
    "\n",
    "# Pseudonymization\n",
    "config = {\n",
    "    \"almighty pseudonymizer\": {\n",
    "        \"type\": \"ng_pseudonymizer\",\n",
    "        \"pubkey_analyst\": \"./analyst.crt\",\n",
    "        \"pubkey_depseudo\": \"./depseudo.crt\",\n",
    "        \"regex_mapping\": \"../../../../../examples/exampledata/rules/pseudonymizer/regex_mapping.yml\",\n",
    "        \"hash_salt\": \"a_secret_tasty_ingredient\",\n",
    "        \"outputs\": [\n",
    "            {\"opensearch\": \"pseudonyms\"}\n",
    "        ],\n",
    "        \"rules\": [\n",
    "            {\n",
    "                \"filter\": \"*\",\n",
    "                \"pseudonymizer\": {\n",
    "                    \"mapping\": {\n",
    "                        \"user.name\": \"RE_WHOLE_FIELD\",\n",
    "                    }\n",
    "                }\n",
    "            }\n",
    "        ],\n",
    "        \"max_cached_pseudonyms\": 1000000\n",
    "    }\n",
    "}\n",
    "processor = Factory.create(config)\n",
    "processor.setup()\n",
    "processor.process(event)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "607bbe20",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(f\"Event after processing: {json.dumps(event.data, indent=2)}\")\n",
    "print(f\"{len(event.extra_data)=}\")\n",
    "print(f\"Event extra data: {json.dumps(event.extra_data[0].data, indent=2)}\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "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": "3.12.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}