Data workflow guides

Check a legacy CSV encoding before an inventory import

·

A warehouse import can reject a supplier’s CSV even when its separators look right. Start with the original saved bytes. Identify what the bytes prove, compare the decoded fields with the supplier’s export, and reopen the converted copy before importing.

CSV Encoding Detector

Inspect the original file bytes

Open the example file or choose Load example. Use Semicolon as the source separator. Inspect with Automatic when verified first. The example has no byte-order mark. Its high bytes do not form valid UTF-8, so choose a source encoding to preview it.

Select windows-1252 and inspect again. Read 3 records with 3 fields per record. The second record contains identifier 001, a quoted note beginning Café, €2 with retour on the next line, and the exact amount 1.2300. The note stays in one field.

ToolMellow Windows-1252 orders sampleWHATWG Encoding StandardRFC 4180, CSV records and quoting

Use the euro sign to compare decodings

The example stores the euro sign as hexadecimal 80. In Windows-1252 this byte produces €. ISO-8859-1 assigns it a control code. Try ISO-8859-1 and inspect again to see the control-code diagnostic, then return to windows-1252.

ASCII-compatible files can have several equally valid interpretations. A byte-order mark gives stronger evidence for Unicode, and strict decoding checks the bytes after the mark. A file without a mark may need the exporting application’s encoding setting.

WHATWG Encoding StandardUnicode FAQ on byte-order marks

Reopen the UTF-8 copy with the same separator

Download UTF-8 CSV, open the downloaded file and inspect it with Automatic when verified and Semicolon. The strict UTF-8 check succeeds. Verify 001, the two-line note and 1.2300 before importing. The third record’s empty amount stays empty.

Choose text import for order identifiers if your spreadsheet would display 001 as 1. If validation still reports extra fields or missing quotes, check the original separator and syntax. An encoding conversion preserves the original record text after the leading signature.

RFC 4180, CSV records and quotingUnicode FAQ on byte-order marks

Common questions

Why inspect a file rather than pasted text?

Pasted text has already been decoded. Start with the saved file to inspect its original bytes, byte order and possible legacy decoding.

Can I recover text already saved as a replacement character?

A replacement character records a prior decoding error without preserving the missing character. Recover the original export if you need that text.

Sources

Put it into practice.

Related guides