CSV Delimiters Explained: Comma, Tab, Semicolon, and Pipe

A CSV file is just plain text, but text alone doesn't tell a program where one column ends and the next begins. That job belongs to the delimiter — the single character that splits each row into fields. The name "CSV" stands for comma-separated values, yet in the real world the separating character is often not a comma at all. Understanding delimiters is the key to opening any delimited file without scrambled columns.

What a delimiter is

Imagine a single line from a spreadsheet export:

Ava,Chen,Phoenix,AZ

A parser reads that line left to right and, every time it hits the delimiter (here a comma), it ends the current field and starts a new one. The result is four values: Ava, Chen, Phoenix, AZ. Change the delimiter and the same characters split differently. So before any tool can show you a clean grid, it has to know which character to break on.

The common delimiters

Four separators cover the vast majority of delimited files you'll encounter, and each tends to travel with its own file extension.

Delimiter	Name	Typical extension	Where it comes from
,	Comma	.csv	The default for English-locale exports, spreadsheets, and most databases
\t (tab)	Tab	.tsv, .tab	Data exports, bioinformatics, anything where values may contain commas
;	Semicolon	.csv	European spreadsheet exports where comma is the decimal separator
|	Pipe	.psv	Database dumps, log files, and feeds where commas and tabs both appear in the data

A few things worth noting about that table. The .csv extension shows up twice because both comma- and semicolon-separated files are commonly saved as .csv — the extension tells you nothing reliable about the actual separator inside. Tab-separated files are common enough to have earned their own format; there's more on that in CSV vs TSV. And pipe-delimited files exist precisely because the pipe character almost never appears in ordinary data, which makes it a safe separator.

Why semicolons appear in European files

This one trips people up constantly. In the United States and the UK, 1,000.50 means one thousand and a half — comma groups the thousands, period marks the decimal. But in much of Europe the roles are swapped: the same number is written 1.000,50, with the comma as the decimal separator.

If a tool used a comma to separate fields in that locale, the number 1,5 would be ambiguous — is it one field (one-and-a-half) or two fields (1 and 5)? To sidestep the collision, spreadsheet software in those regions exports CSVs using a semicolon as the field separator. The file still ends in .csv, but a parser expecting commas will see a single garbled column. The fix is simply to parse on the semicolon instead, covered in detail in opening semicolon CSV files.

Quoting: when a field contains the delimiter

Delimiters create a problem the moment your actual data contains the same character. Consider a company name like Smith, Jones & Co.. If the comma inside the name were treated as a separator, that one field would split into two and every column after it would shift.

The CSV convention solves this with quoting. A field that contains the delimiter (or a line break, or a quote character) is wrapped in double quotes:

"Smith, Jones & Co.",Phoenix,AZ

Now the parser knows that everything between the quotes is a single value, comma included. If the field itself contains a double quote, that quote is doubled ("") so it isn't mistaken for the end of the field. This is why robust CSV parsing is more than "split on the comma" — the quoting rules have to be respected, or quoted fields with embedded delimiters break the whole row.

How auto-detection works

Because the file extension lies and the separator could be any of four characters, a good viewer figures out the delimiter for you. The approach is straightforward: scan the first chunk of the file, try each candidate separator (comma, tab, semicolon, pipe), and see which one produces the most consistent column count across rows. A real comma-separated file yields the same number of columns on every line when split on commas; splitting it on semicolons would give one big column. The separator that lines up cleanly is the winner.

CEESVEE does exactly this. When you open a file it auto-detects both the delimiter and the encoding (UTF-8, UTF-16 LE/BE, or Windows-1252) and handles the byte-order mark correctly, so the grid is usually right on the first try. When you need to take over — an unusual separator, or a file the heuristic guesses wrong — there's a manual override with a custom-delimiter option, and the grid re-parses instantly so you can confirm the columns look right.

Changing the delimiter on the way out

Reading is only half the round-trip. Sometimes you receive a semicolon file but a downstream system demands true commas, or you need to convert a CSV into a tab- or pipe-delimited feed. On Save or Save As, CEESVEE lets you pick the export delimiter explicitly, along with the encoding, quoting style, line endings (LF or CRLF), and whether to write a BOM. That makes converting between separator styles a deliberate, repeatable step rather than a guessing game — see how to change a CSV's delimiter for a walkthrough.

The bottom line

A delimiter is the one character that turns a line of text into columns, and "CSV" is a loose label that hides commas, tabs, semicolons, and pipes alike. Once you know what to look for — the right separator on read, proper quoting for fields that contain it, and an explicit delimiter on export — delimited files stop being mysterious.

Download CEESVEE for free and let it detect the delimiter for you, so your columns line up the moment the file opens.