CSV 转 JSON 转换器

The CSV rules that trip people up

CSV looks simple but has a few sharp edges, all defined by RFC 4180:

1. Values containing a comma must be quoted — Smith, John breaks without quotes
2. Values containing a quote must be quoted, and the quote doubled — "He said ""hi"""
3. Values containing a newline must be quoted
4. The first row is the header by convention

This tool implements all four, so pasted CSV that looks broken in a text editor still parses correctly.

Example

name,role,year
Ada,Engineer,1815
Alan,Researcher,1912

becomes

[
  { "name": "Ada", "role": "Engineer", "year": "1815" },
  { "name": "Alan", "role": "Researcher", "year": "1912" }
]

Types are not inferred

Every value comes out as a string, because CSV has no type information. If you need numbers, booleans or nulls, cast them in code afterwards, or use a tool that supports typed columns.

For a quick conversion in JavaScript:

const rows = data.map(r => ({
  name: r.name,
  year: Number(r.year),
  active: r.active === "true",
}));

Command line

For large files, jq handles CSV with the @csv operator:

# CSV to JSON
cat data.csv | jq -R -s 'split("\n") | map(split(",")) | transpose | ...'

Or use csvjson if you have it installed — it handles types and headers properly.

Related

Going the other way? The JSON to CSV converter flattens an array of objects into spreadsheet-ready rows.