Common CSV Formatting Errors
A CSV file that opens as one giant column, shows garbage characters, or shifts every value one cell left is not corrupt — it is one of five well-understood problems, each with a specific fix.
1. Quoting errors — values split at the wrong comma
CSV has one rule that does all the heavy lifting: a value containing the delimiter, a double quote, or a newline must be wrapped in double quotes, and quotes inside it are doubled. The classic breakage:
name,note Roberta,Paid in full, thanks Oliver,He said "hello" today
Row 1 has three fields under two headers — thanks spills into a phantom column and every value after it shifts. Row 2's unescaped quotes will be eaten or misread by most parsers. Correct output quotes both notes: "Paid in full, thanks" and "He said ""hello"" today". If you received the broken file, there is no fully automatic rescue — the information about where fields end is genuinely missing — but if you are producing CSV, always generate it with a real CSV writer (our converters quote per RFC 4180) rather than joining strings with commas by hand.
2. Delimiter mismatches — everything in one column
If the whole file lands in a single column, the parser guessed the wrong separator. European Excel exports use semicolons (because commas are decimal points there); database exports often use tabs or pipes. Open the file's first line in a text editor and look at what actually separates the values, then open it with that delimiter — our CSV Viewer auto-detects comma, semicolon, tab and pipe, and the TXT to CSV converter lets you set it manually and re-export as standard comma CSV.
3. Encoding problems — accents turn into garbage
Characters like é, ü or — appearing as é or � mean the file's encoding and the reader's encoding disagree — usually a Windows-1252 or Latin-1 file read as UTF-8, or the reverse. The reliable fix is to re-save the file as UTF-8 (any good text editor can), ideally with a byte-order mark (BOM) if Excel is the consumer, since Excel uses the BOM to detect UTF-8. The data is almost never lost; it is just being read with the wrong alphabet.
4. Line-ending problems — everything on one line, or double-spaced rows
Windows ends lines with CRLF, Unix with LF. Most modern parsers handle both, but a file with mixed endings — common when files are edited by hand or concatenated — produces merged or phantom empty rows. Normalising line endings in a text editor, or re-exporting through a converter, fixes it.
5. Dirty data — duplicates, blank rows, stray spaces
The file parses fine but is still wrong: the same row twice, empty rows scattered through it, values padded with spaces (" Roberta") that break lookups and joins. These are cleaning problems, not format problems. The CSV Cleaner removes duplicate and empty rows, trims every cell and drops empty columns — and shows you a before/after report so you can see exactly what changed before downloading.
A 30-second diagnosis routine
One column → delimiter. Shifted values → quoting. Garbage accents → encoding. Merged or phantom rows → line endings. Right shape, wrong data → cleaning. Match the symptom to the cause before reaching for a fix, and the fix is usually one step.