How does it decide which delimiter is in use?

It counts how many times each candidate delimiter (comma, semicolon, tab, pipe) appears outside quoted fields on every line, then picks whichever candidate has the same count on the most lines, since a real column separator should appear the same number of times on every row.

What happens to fields that already contain a comma?

When re-serializing to comma-delimited output, any field containing a comma, a double quote, or a newline is wrapped in double quotes, with internal double quotes doubled, per standard CSV quoting rules.

What if my data uses two different delimiters in different places?

The detector picks whichever single delimiter is most consistent across lines; if no candidate is consistent enough, it reports that no delimiter could be reliably detected instead of guessing.

Can it detect delimiters other than comma, semicolon, tab, and pipe?

No, detection is limited to those four candidates, which cover the overwhelming majority of real-world CSV/TSV-like exports.

CSV Delimiter Detector - Detect & Normalize Comma, Semicolon, Tab, Pipe

Overview

Introduction

CSV exports don't always use commas. Excel's regional settings, European database exports, and log files pasted from a terminal frequently use semicolons, tabs, or pipes instead, and pasting that into a comma-only CSV tool silently produces one giant unparsed column.

This tool figures out which delimiter a pasted block of text is actually using and rewrites it as standard, comma-delimited CSV so downstream tools that expect commas work correctly.

What Is CSV Delimiter Detector and Normalizer?

A delimiter detector and normalizer for CSV-like text: it inspects a pasted block of rows, works out which of comma, semicolon, tab, or pipe is being used as the column separator, and outputs the same data re-written with commas.

It runs entirely client-side as part of this site's String Tools collection, so nothing you paste is ever uploaded to a server.

How CSV Delimiter Detector and Normalizer Works

For each candidate delimiter, the tool counts how many times it appears on every line while ignoring anything inside double-quoted fields (so a comma inside a quoted address doesn't get miscounted as a column separator). The delimiter whose count matches most consistently across all lines, and is non-zero, is reported as detected.

The detected delimiter is then used to parse the text into rows with this repo's shared delimited-text parser (which honors RFC 4180 quoting, including quoted fields that span multiple lines), and the rows are re-serialized as comma-delimited CSV, re-quoting any field that itself contains a comma, quote, or newline.

When To Use CSV Delimiter Detector and Normalizer

Use it right after pasting an export from a tool or region that doesn't default to commas, a European Excel export using semicolons, a database dump using pipes, or a log file using tabs, before feeding the data into a comma-only CSV parser.

It's a fast way to get consistent, comma-delimited CSV without opening a spreadsheet application or writing a one-off script just to check.

Often used alongside CSV to JSON Converter, CSV Joiner & Column Splicer and CSV to TOON.

Features

Advantages

Detects the delimiter automatically instead of requiring you to specify it up front.
Ignores delimiter characters that appear inside quoted fields, so it doesn't get confused by a comma inside an address or a quoted number.
Re-quotes fields correctly on output, so values containing commas, quotes, or newlines survive the round trip to comma-delimited CSV.

Limitations

Detection is heuristic, based on consistency of counts across lines, and can be wrong on very short (one or two row) or highly irregular inputs.
Only comma, semicolon, tab, and pipe are considered; any other single-character delimiter won't be detected.

Examples

Detecting and normalizing a semicolon-delimited CSV

Input

name;age;city
Ada;30;London
Grace;40;Paris

Output

name,age,city
Ada,30,London
Grace,40,Paris

Semicolon appears exactly twice on every line, so it's detected as the delimiter and the output is re-written with commas.

Best Practices & Notes

Best Practices

Check the reported delimiter before trusting the normalized output on unusual or very short inputs, where the heuristic has less data to work with.
Run the CSV Encoding Detector first if the source file might not be UTF-8, since delimiter detection works on already-decoded text.
Feed the normalized comma-delimited output straight into Convert CSV to JSON or the CSV Joiner & Column Splicer, both of which expect commas.

Developer Notes

Delimiter counting scans each line character by character, toggling an `inQuotes` flag on unescaped double quotes so delimiters inside quoted fields are never counted; consistency is scored as the fraction of lines whose count matches the first line's count for that candidate, and the highest-scoring, non-zero candidate wins. Parsing and re-serialization reuse this repo's shared `parseDelimited`/`serializeCsv` helpers rather than a second hand-rolled parser.

CSV Delimiter Detector and Normalizer Use Cases

Normalizing a semicolon-delimited export from a European locale Excel before importing it elsewhere
Converting a tab-separated log or database dump into standard CSV
Cleaning up pipe-delimited data pasted from a legacy system before running it through a comma-only tool

Common Mistakes

Assuming pasted data is comma-delimited just because it's called a "CSV" file; many real-world exports use semicolons or tabs instead.
Not noticing that a comma inside a quoted field was being miscounted as a delimiter by a naive detector; this tool specifically guards against that by ignoring delimiters inside quotes.

Tips

If detection fails, check whether the pasted text actually has consistent columns at all; ragged rows with a varying number of fields will defeat the consistency heuristic.
Paste only the data rows plus header, without extra leading or trailing blank lines, for the most reliable detection.

CSV Delimiter Detector and Normalizer