csv

Parse CSV and TSV text into rows and fields

csv parses CSV-style delimited text into rows and fields.

With the default comma delimiter, parseDocument() implements RFC 4180 record and quoted-field parsing, including embedded CRLF, doubled-quote escaping, consistent field counts, and explicit parse errors for malformed input.

It also supports a few explicit extensions: alternate delimiters, optional empty-line skipping, and LF-only line endings.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18

const std = @import("std");
const zigcli = @import("zigcli");
const csv = zigcli.csv;

const allocator = std.heap.page_allocator;
const input =
    "name,age,city\r\n" ++
    "alice,30,beijing\r\n" ++
    "bob,25,\"new york\"\r\n";

var document = try csv.parseDocument(allocator, input, .{
    .delimiter = ',',
});
defer document.deinit(allocator);

for (document.rows) |row| {
    std.debug.print("{s}\n", .{row[0]});
}

API

parseLine(allocator, line, delimiter)

Parse one line and return a newly allocated slice of fields.

parseDocument(allocator, input, options)

Parse a whole input buffer and return a Document:

  • rows: parsed rows
  • num_cols: column count inferred from the first parsed row

Malformed input returns a parse error such as error.InvalidQuotedField, error.InvalidRecordTerminator, error.InconsistentFieldCount, or error.UnclosedQuotedField.

Document owns the row and field storage. Call document.deinit(allocator) when you are done with it.

Last modified March 22, 2026: chore: bump to 0.5.0 (8c7d27e)