CSV validator
Check CSV quoting, column counts and required values with original source line references. Your file stays in your browser.
Selected CSV rules
Invalid CSV
Complete data rows
2
Source lines
4
Expected columns
3
Errors
1
Checks apply only to the selected file and column rules.
Header summary
sku | name | price
Validation errors
| Source line | Data row | Column | Error |
|---|---|---|---|
| 4 | 2 | 1 | This required field is empty or contains only whitespace. |
All errors are shown.
JSON report with rules and error locations. Your CSV stays unchanged.
Invalid CSV. One required value is missing.
Choose the CSV rules
This CSV validator checks your local file for broken quoting, uneven columns and the required fields you select. Paste text or choose a file.
Select the delimiter used by the source. Comma, semicolon, tab and pipe are supported. The encoding setting controls how file bytes become text. It has no effect on pasted text.
Double quotes enclose a field that contains a delimiter or line break. Inside that field, two consecutive quotes mean one literal quote. No quoting treats quote characters as ordinary text.
The quoting rules follow RFC 4180’s CSV format description. This tool also accepts LF and CR record endings, alternative delimiters and Unicode text in the selected encoding.
Choose whether the first record is a header. An empty Expected columns field takes the first complete record’s width. Enter a number to check against a known import width.
Required headers match names exactly, including case and spaces. Non-empty fields use those names with a header, or one-based positions without one. Whitespace-only values fail that check.
These requirements come from you. A store’s import template can require different headers, value types or permitted values.
Read the validation errors
Each validation error gives the original source line and parsed data row. Column is a field position, counted from one.
A quoted field can span several source lines while remaining one data row. In the loaded example, the first product spans lines 2 and 3. The second product is data row 2 on source line 4.
A field-count error compares the entire record with the expected width. A trailing delimiter creates an empty final field, so it still counts as a column.
A quote error points to the failing field or its opening quote. Parsing stops at that syntax error because later row boundaries cannot be trusted. The data count includes only complete records before it.
A missing header appears at the header record, while an empty required value points to the relevant field. A required header repeated twice also fails because the field name is ambiguous.
The tool checks up to 10,000 records including the header and 100 columns. The report contains the first 100 errors and the total found. A limit failure means validation is incomplete.
A valid result confirms the selected structure rules. It cannot establish that a price is correct, a product exists or an import will succeed on a particular platform.
Correct and check the original
The fictional supplier export opens with two product records and a required sku field. Its result is Invalid CSV because the second product’s SKU is empty.
- Find source line 4, data row 2, column 1 in the error list.
- For this fictional example, type
CUP-002before the comma at the start of that line. For your own file, use the identifier from your catalogue. - Keep the same rules and check again. The example becomes Valid CSV with two complete data rows and no errors.
Keep your source file and correct a working copy. Selecting a file or downloading a report never overwrites the original.
The JSON report contains the selected rules, header names and error locations. It contains no converted or automatically repaired CSV. Header names can contain private information, so handle the report with the same care as your source.
For delimiter changes, field mapping or another output format after validation, use the CSV converter.