{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import re\n",
    "import pandas as pd\n",
    "\n",
    "# Function to extract log entries with specified pattern and check for 429 errors\n",
    "def extract_429_error_log_entries(file_path):\n",
    "    with open(file_path, 'r') as file:\n",
    "        log_entries = file.readlines()\n",
    "    \n",
    "    log_data = []\n",
    "    pattern = re.compile(r'(\\d{4}-\\d{2}-\\d{2} \\d{2}:\\d{2}:\\d{2},\\d{3}) - ERROR - tool_handler Tool call failed:')\n",
    "    \n",
    "    for i in range(len(log_entries)):\n",
    "        entry = log_entries[i]\n",
    "        match = pattern.match(entry)\n",
    "        if match:\n",
    "            # Check within the next 100 lines for a 429 error\n",
    "            for j in range(i + 1, min(i + 100, len(log_entries))):\n",
    "                if '<h1>429</h1>' in log_entries[j]:\n",
    "                    next_timestamp_match = re.match(r'\\d{4}-\\d{2}-\\d{2} \\d{2}:\\d{2}:\\d{2},\\d{3}', log_entries[j + 1]) if j + 1 < len(log_entries) else None\n",
    "                    next_timestamp = next_timestamp_match.group() if next_timestamp_match else None\n",
    "                    log_data.append({\n",
    "                        'Timestamp': match.group(1),\n",
    "                        'Log Entry': entry.strip(),\n",
    "                        '429 Error Entry': log_entries[j].strip(),\n",
    "                    })\n",
    "                    break\n",
    "    \n",
    "    return pd.DataFrame(log_data)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extract the log entries matching the specified pattern and containing 429 errors\n",
    "file_path = 'paste_v2.txt'\n",
    "error_429_log_df = extract_429_error_log_entries(file_path)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Timestamp</th>\n",
       "      <th>Log Entry</th>\n",
       "      <th>429 Error Entry</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2024-06-25 06:39:01,261</td>\n",
       "      <td>2024-06-25 06:39:01,261 - ERROR - tool_handler...</td>\n",
       "      <td>&lt;h1&gt;429&lt;/h1&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2024-06-25 06:39:03,968</td>\n",
       "      <td>2024-06-25 06:39:03,968 - ERROR - tool_handler...</td>\n",
       "      <td>&lt;h1&gt;429&lt;/h1&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2024-06-25 06:39:06,904</td>\n",
       "      <td>2024-06-25 06:39:06,904 - ERROR - tool_handler...</td>\n",
       "      <td>&lt;h1&gt;429&lt;/h1&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2024-06-25 07:52:06,830</td>\n",
       "      <td>2024-06-25 07:52:06,830 - ERROR - tool_handler...</td>\n",
       "      <td>&lt;h1&gt;429&lt;/h1&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2024-06-25 07:52:09,671</td>\n",
       "      <td>2024-06-25 07:52:09,671 - ERROR - tool_handler...</td>\n",
       "      <td>&lt;h1&gt;429&lt;/h1&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2024-06-25 07:52:33,216</td>\n",
       "      <td>2024-06-25 07:52:33,216 - ERROR - tool_handler...</td>\n",
       "      <td>&lt;h1&gt;429&lt;/h1&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2024-06-25 07:52:35,952</td>\n",
       "      <td>2024-06-25 07:52:35,952 - ERROR - tool_handler...</td>\n",
       "      <td>&lt;h1&gt;429&lt;/h1&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2024-06-25 07:54:39,687</td>\n",
       "      <td>2024-06-25 07:54:39,687 - ERROR - tool_handler...</td>\n",
       "      <td>&lt;h1&gt;429&lt;/h1&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2024-06-25 07:54:42,478</td>\n",
       "      <td>2024-06-25 07:54:42,478 - ERROR - tool_handler...</td>\n",
       "      <td>&lt;h1&gt;429&lt;/h1&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2024-06-25 07:57:06,539</td>\n",
       "      <td>2024-06-25 07:57:06,539 - ERROR - tool_handler...</td>\n",
       "      <td>&lt;h1&gt;429&lt;/h1&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>2024-06-25 08:14:07,060</td>\n",
       "      <td>2024-06-25 08:14:07,060 - ERROR - tool_handler...</td>\n",
       "      <td>&lt;h1&gt;429&lt;/h1&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>2024-06-25 08:14:10,071</td>\n",
       "      <td>2024-06-25 08:14:10,071 - ERROR - tool_handler...</td>\n",
       "      <td>&lt;h1&gt;429&lt;/h1&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>2024-06-25 08:14:12,738</td>\n",
       "      <td>2024-06-25 08:14:12,738 - ERROR - tool_handler...</td>\n",
       "      <td>&lt;h1&gt;429&lt;/h1&gt;</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  Timestamp  \\\n",
       "0   2024-06-25 06:39:01,261   \n",
       "1   2024-06-25 06:39:03,968   \n",
       "2   2024-06-25 06:39:06,904   \n",
       "3   2024-06-25 07:52:06,830   \n",
       "4   2024-06-25 07:52:09,671   \n",
       "5   2024-06-25 07:52:33,216   \n",
       "6   2024-06-25 07:52:35,952   \n",
       "7   2024-06-25 07:54:39,687   \n",
       "8   2024-06-25 07:54:42,478   \n",
       "9   2024-06-25 07:57:06,539   \n",
       "10  2024-06-25 08:14:07,060   \n",
       "11  2024-06-25 08:14:10,071   \n",
       "12  2024-06-25 08:14:12,738   \n",
       "\n",
       "                                            Log Entry 429 Error Entry  \n",
       "0   2024-06-25 06:39:01,261 - ERROR - tool_handler...    <h1>429</h1>  \n",
       "1   2024-06-25 06:39:03,968 - ERROR - tool_handler...    <h1>429</h1>  \n",
       "2   2024-06-25 06:39:06,904 - ERROR - tool_handler...    <h1>429</h1>  \n",
       "3   2024-06-25 07:52:06,830 - ERROR - tool_handler...    <h1>429</h1>  \n",
       "4   2024-06-25 07:52:09,671 - ERROR - tool_handler...    <h1>429</h1>  \n",
       "5   2024-06-25 07:52:33,216 - ERROR - tool_handler...    <h1>429</h1>  \n",
       "6   2024-06-25 07:52:35,952 - ERROR - tool_handler...    <h1>429</h1>  \n",
       "7   2024-06-25 07:54:39,687 - ERROR - tool_handler...    <h1>429</h1>  \n",
       "8   2024-06-25 07:54:42,478 - ERROR - tool_handler...    <h1>429</h1>  \n",
       "9   2024-06-25 07:57:06,539 - ERROR - tool_handler...    <h1>429</h1>  \n",
       "10  2024-06-25 08:14:07,060 - ERROR - tool_handler...    <h1>429</h1>  \n",
       "11  2024-06-25 08:14:10,071 - ERROR - tool_handler...    <h1>429</h1>  \n",
       "12  2024-06-25 08:14:12,738 - ERROR - tool_handler...    <h1>429</h1>  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(error_429_log_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}