CSV to JSON Converter

CSV to JSON is a small job that a naive one-liner gets wrong in a dozen ways, so this page is built around a real RFC 4180 parser written out by hand. A quoted cell that contains commas, a cell with a line break in the middle of it, a doubled quote inside a quoted cell, a file that ends CRLF, a file that ends with a bare CR, a UTF-8 byte order mark on the front, a trailing line ending, and a header row whose names are not unique are all handled as cases with names — not as things that happen to work.

The parser is deliberately honest about the parts that are guesses. A double quote in the middle of an unquoted field is reported as an anomaly with its record, field and character position instead of being swallowed, because that is exactly how a converter loses a column. A row with fewer fields than the header has its missing cells written as JSON null rather than as an empty string, since an empty cell is something the CSV said and a missing cell is something it did not. A row with more fields keeps its extra values in a named array rather than dropping them. Duplicate header names are renamed deterministically and every rename is listed on screen.

No type is inferred. Every value in the output is a JSON string, so 00123 keeps its leading zeros, TRUE stays TRUE rather than becoming true, and a date in whatever format the export wrote stays in that format. That is a decision, not a limitation being hidden: a spreadsheet guesses types on open and this page does not, and it says so in the same font as the rest of the copy.

Input is capped before any parsing work begins — 5,242,880 characters of pasted text, 5 MB of opened file, 20,000 rows, 512 columns per row and 100,000 characters in a single cell — and each cap is refused with the real numbers rather than silently truncating. Your CSV is read in this tab. Nothing is uploaded, and no request is made with your data at all.