Challenge: CSV parser

The CSV format (Comma-Separated Values) is everywhere: exports from Excel, database dumps, output from command-line tools. Parsing it is a small challenge that combines split, map, destructuring, and object construction.

The format

A minimal CSV is a multi-line string: the first row is the header (column names), then one row per record with fields separated by commas.

nome,eta,citta
Alice,30,Roma
Bob,25,Milano

The classic goal is to turn it into an array of objects:

[
  { nome: 'Alice', eta: '30', citta: 'Roma' },
  { nome: 'Bob', eta: '25', citta: 'Milano' },
];

The 3-step recipe

1. Split the string by lines: csv.split('\n')
2. Extract the header row: the first line, split by comma
3. Map each remaining line to an object, pairing each field with the matching header

function parseCSV(csv) {
  const righe = csv.split('\n');
  const intestazione = righe[0].split(',');
  return righe.slice(1).map((riga) => {
    const campi = riga.split(',');
    const record = {};
    intestazione.forEach((nome, i) => {
      record[nome] = campi[i];
    });
    return record;
  });
}

More modern alternatives:

const record = Object.fromEntries(intestazione.map((nome, i) => [nome, campi[i]]));

Converting types

Parsed fields are always strings. If you know that some columns are numbers:

record.eta = Number(record.eta);

Or write a small "schema":

const numeriche = new Set(['eta', 'prezzo']);
intestazione.forEach((nome, i) => {
  const v = campi[i];
  record[nome] = numeriche.has(nome) ? Number(v) : v;
});

Try it

Review exercise