CrazeAnalytix/test-csv.js

import fs from 'fs';
import https from 'https';

const url = "https://www.dropbox.com/scl/fi/b9zxn4z5i7sxwfakk5g5y/Amazon-Sell-Out-2023-2025.csv?rlkey=uoto6v0mm99py8nszy8ldtez8&dl=1";

https.get(url, (res) => {
  if (res.statusCode >= 300 && res.statusCode < 400 && res.headers.location) {
    https.get(res.headers.location, (res2) => {
      processData(res2);
    });
  } else {
    processData(res);
  }
}).on('error', (e) => {
  console.error(e);
});

function processData(stream) {
  let data = '';
  stream.on('data', chunk => {
    data += chunk.toString('utf8');
  });
  stream.on('end', () => {
    analyze(data);
  });
}

function analyze(data) {
  const lines = data.split('\n');
  console.log("Total Lines in CSV:", lines.length);

  let maxYear = 0;
  let maxWeek = 0;
  let minYear = 2030;
  let minWeek = 52;
  let legendsRows = 0;
  let legendsMaxYear = 0;
  let legendsMaxWeek = 0;

  for (let i = 1; i < lines.length; i++) {
    const row = lines[i].split(',');
    if (row.length < 5) continue;
    const year = parseInt(row[3]); // Based on app processor logic it's mostly col 3 or 1
    const week = parseInt(row[4]); // Based on app processor

    // Use regex fallback if parsing fails
    const matchedYear = parseInt(row.find(r => r.startsWith('202')) || '0');
    const matchedWeek = parseInt(row.find(r => r.match(/^[0-9]{1,2}$/)) || '0');

    const y = year || matchedYear || 0;
    const w = week || matchedWeek || 0;

    if (y > maxYear) { maxYear = y; maxWeek = w; }
    else if (y === maxYear && w > maxWeek) { maxWeek = w; }

    if (y < minYear && y > 2000) { minYear = y; minWeek = w; }

    if (lines[i].toLowerCase().includes('legends')) {
      legendsRows++;
      if (y > legendsMaxYear) { legendsMaxYear = y; legendsMaxWeek = w; }
      else if (y === legendsMaxYear && w > legendsMaxWeek) { legendsMaxWeek = w; }
    }
  }
  console.log(`Global -> Min Date: Year ${minYear}, Week ${minWeek} | Max Date: Year ${maxYear}, Week ${maxWeek}`);
  console.log(`LEGENDS stats -> Total rows: ${legendsRows}, Max Date: Year ${legendsMaxYear}, Week ${legendsMaxWeek}`);
}