{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "def csv_to_dataframe(file_path):\n",
    "\n",
    "    import pandas as pd\n",
    "    import os\n",
    "    from io import StringIO\n",
    "    \n",
    "    # open the file \n",
    "    \n",
    "    file_reader = open(file_path)\n",
    "\n",
    "    # create the empty dataframe\n",
    "    \n",
    "    headers = file_reader.readline()\n",
    "    headers_ls = headers.split(',')\n",
    "    df_global = pd.DataFrame(columns=headers_ls)\n",
    "\n",
    "    # prepare the error file\n",
    "    \n",
    "    dir_path = os.path.dirname(file_path)\n",
    "    file_name = os.path.basename(file_path)\n",
    "    file_base, file_ext = os.path.splitext(file_name)\n",
    "    \n",
    "    error_file_noext = os.path.join(dir_path, file_base)\n",
    "    error_file_path = f'{error_file_noext}_error.txt'\n",
    "    #print(error_file_path) \n",
    "    \n",
    "    line_no = 2\n",
    "    error_file = open(error_file_path, 'w+')\n",
    "    error_lines = []\n",
    "    \n",
    "    # read one line at a time\n",
    "    \n",
    "    for line in file_reader:\n",
    "        \n",
    "        # if there are no errors, append it to dataframe\n",
    "\n",
    "        try:\n",
    "            #print(line)\n",
    "            df_line = pd.read_csv(StringIO(line), header=None)  # header none makes it treat 1 line like data\n",
    "\n",
    "            # will error here if number of columns in df_line doesn't match number in df_global\n",
    "            df_line.columns = df_global.columns\n",
    "            #df_line.info()    \n",
    "            df_global = df_global.append(df_line)\n",
    "\n",
    "        # if there is an error, store it and output it\n",
    "        \n",
    "        except Exception as ex:\n",
    "            #print(f'{line_no} : {ex}')\n",
    "            \n",
    "            error_lines.append(line)\n",
    "            error_file.write(line)\n",
    "\n",
    "        line_no += 1\n",
    "        #if line_no > 0:\n",
    "        #    break\n",
    "    \n",
    "    return df_global, error_lines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 1525 entries, 0 to 0\n",
      "Data columns (total 10 columns):\n",
      "               1525 non-null object\n",
      "business_id    1525 non-null object\n",
      "cool           1525 non-null object\n",
      "date           1525 non-null object\n",
      "funny          1525 non-null object\n",
      "review_id      1525 non-null object\n",
      "stars          1525 non-null object\n",
      "text           1525 non-null object\n",
      "useful         1525 non-null object\n",
      "user_id\n",
      "       1525 non-null object\n",
      "dtypes: object(10)\n",
      "memory usage: 131.1+ KB\n"
     ]
    }
   ],
   "source": [
    "# call the function and inspect the DataFrame\n",
    "\n",
    "df_final, errors = csv_to_dataframe('Data/Yelp_Reviews_corrupt.csv')\n",
    "df_final.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of errors found: 4583\n"
     ]
    }
   ],
   "source": [
    "# check if there were errors in the file\n",
    "\n",
    "msg = f'number of errors found: {len(errors)}'\n",
    "print(msg)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}