{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Processing Notebook\n",
    "\n",
    "The purpose of this notebook is to help me process the RAW S3 data into the right format to upload to Hugging Face."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import dependencies\n",
    "import os\n",
    "import json\n",
    "from datetime import datetime"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make directory called `data/faa` if it doesn't already exist.\n",
    "if not os.path.exists('data/faa'):\n",
    "\tos.makedirs('data/faa')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wrote 41164 records to data/faa/faa_xml_data.txt\n",
      "Wrote 41164 records to data/faa/faa_xml_data.jsonl\n"
     ]
    }
   ],
   "source": [
    "# For each file in `s3_data` directory\n",
    "xml_data_strings = []\n",
    "for file in os.listdir('s3_data'):\n",
    "\t# If file doesn't end in `.txt`, continue\n",
    "\tif not file.endswith('.txt'):\n",
    "\t\tcontinue\n",
    "\t# Open the file in read mode\n",
    "\twith open(f's3_data/{file}', 'r') as f:\n",
    "\t\t# Read the file content\n",
    "\t\tcontent = f.read()\n",
    "\n",
    "\t\t# Split the content by the separator\n",
    "\t\tparts = content.split('<><><><><>')\n",
    "\n",
    "\t\txml_data = ''\n",
    "\t\t# Get the XML data as a string\n",
    "\t\tif len(parts) == 1:\n",
    "\t\t\txml_data_strings.append(parts[0].split('---')[1].strip())\n",
    "\t\telse:\n",
    "\t\t\tfor part in parts:\n",
    "\t\t\t\txml_data_strings.append(part.split('---')[1].strip())\n",
    "\n",
    "# Delete any duplicate strings in the xml_data_strings list\n",
    "xml_data_strings = list(set(xml_data_strings))\n",
    "\n",
    "# Sort the XML data strings by the `<AIRPORT_STATUS_INFORMATION><Update_Time>Wed Jul 10 17:18:22 2024 GMT</Update_Time>` date\n",
    "def extract_update_time(xml_string):\n",
    "\tstart = xml_string.find(\"<Update_Time>\") + len(\"<Update_Time>\")\n",
    "\tend = xml_string.find(\"</Update_Time>\")\n",
    "\tdate_str = xml_string[start:end].strip()\n",
    "\treturn datetime.strptime(date_str, \"%a %b %d %H:%M:%S %Y %Z\")\n",
    "xml_data_strings.sort(key=extract_update_time)\n",
    "\n",
    "# Write the XML data to a TXT file (one XML record per line)\n",
    "with open('data/faa/faa_xml_data.txt', 'w') as f:\n",
    "\tfor xml_data in xml_data_strings:\n",
    "\t\tf.write(xml_data + '\\n')\n",
    "\n",
    "# Write the XML data to a JSONL file (JSON Lines format - one JSON object per line)\n",
    "# This format is well-supported by Hugging Face datasets\n",
    "with open('data/faa/faa_xml_data.jsonl', 'w') as f:\n",
    "\tfor xml_data in xml_data_strings:\n",
    "\t\t# Extract the update time for convenience\n",
    "\t\tupdate_time = extract_update_time(xml_data).isoformat()\n",
    "\t\t# Write each record as a JSON object on its own line\n",
    "\t\tjson_record = {\n",
    "\t\t\t\"update_time\": update_time + \"Z\",\n",
    "\t\t\t\"xml\": xml_data\n",
    "\t\t}\n",
    "\t\tf.write(json.dumps(json_record) + '\\n')\n",
    "\n",
    "print(f\"Wrote {len(xml_data_strings)} records to data/faa/faa_xml_data.txt\")\n",
    "print(f\"Wrote {len(xml_data_strings)} records to data/faa/faa_xml_data.jsonl\")\n",
    "\n",
    "# Delete the xml_data_strings list to free up memory\n",
    "del xml_data_strings"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "3.13.0",
   "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": "3.13.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}