CSV STRUCTURE / A PRACTICAL GUIDE

Why do CSV columns go wrong?

A table can look fine in a text editor and still import into one giant column—or shift values into the wrong fields. Start with the delimiter and quoting rules before changing the data.

When everything lands in one column

A CSV separates fields with commas. Other exports use semicolons, tabs, or pipes. If the reader expects commas but the file uses semicolons, an entire record can become one value.

customer_id;name;city
00124;Amina;Pune

In CSV Ready, choose Semicolon as the input delimiter. If you want comma-separated output, keep Comma (CSV) as the output delimiter. Auto detection is a useful starting guess; always check that the preview matches your expected columns. A file with only one column may be perfectly intentional.

A comma inside a value needs quotes

A customer called “Lee, Sam” should occupy one name field. Put the whole value inside double quotes so its comma is not treated as a field separator.

customer_id,name,city
00125,"Lee, Sam",Mumbai

If a value itself contains a quote, double that quote inside the quoted field:

00126,"Jo ""JJ"" Rao",Delhi

A quote inside an unquoted field, or text after a closing quote, is ambiguous. CSV Ready stops export and reports the record and source line. Edit the source carefully and inspect again. It will not guess which values belong together.

A record is not always one physical line

A quoted cell can contain a line break:

id,notes
00124,"First line
Second line"

That is two records: the header and one data record. The data record spans two physical lines. Diagnostics distinguish the record number from the source line so you can locate the actual input. The preview and export preserve embedded line breaks.

Short rows and long rows need different decisions

If the first record has three fields but a later record has two, a value may be missing. Check the destination’s requirements before enabling Pad short records with empty cells. Padding adds empty fields at the end; it cannot determine whether a value was missing in the middle.

A record with four fields may contain an unquoted comma or a genuine extra value. CSV Ready blocks export while that mismatch remains. It never discards extra fields to fit the expected width.

Empty records are kept by default and can produce width errors. Remove them only if they have no meaning in your data. Exact deduplication also removes whole data records, so check whether repetition is intentional.

Check headers and the destination’s rules

Turn off First record contains column headers when the file has no header. With it enabled, CSV Ready flags empty and exactly duplicated header names. Those warnings do not prove the destination will reject the file; import rules vary.

A consistent structure does not validate required fields, date formats, allowed values, relationships, or a destination schema. Review those requirements in the receiving app before importing.

Next: Preserve identifiers and review spreadsheet formulas.

Format reference: RFC 4180 describes common CSV quoting and record rules. CSV Ready also accepts LF and CR line endings and selected alternative delimiters.