Count CSV records, not lines on the screen
By Hello Keyboard (AI-assisted)
A CSV file can have more visible lines than records. That matters when you change its separator: splitting the text at every newline can turn one note into two rows, even if the file looks tidy afterward.
The useful check is whether the same records and field values survive, not whether the punctuation stays identical.
Try a note that spans two lines
Open Hello Keyboard Change CSV Separator. Paste this invented sample into Or paste CSV text, keeping the line break inside the note:
item,note,qty
"Brass, small","Line one
Line two; ""check""",2
Steel,No change,3There are four visible lines, but three logical records: the header and two data records. Each record has three fields. The first note contains a line break, a semicolon and the word check surrounded by quotation marks.
In this CSV text, a pair of double quotes inside a quoted field represents one literal quotation mark. The comma in Brass, small belongs to the item name, not to a column boundary.
Convert the separator
Set Current separator to Comma (,) and New separator to Semicolon (;). Select Convert separator.
For this sample, the preview is:
item;note;qty
Brass, small;"Line one
Line two; ""check""";2
Steel;No change;3The status reports Converted 3 rows. Here that count includes the header, so it does not mean three inventory items.
Notice two different changes:
• Brass, small no longer needs surrounding quotes when semicolons separate fields. Its comma is still part of the value.
• The note remains quoted. Its internal semicolon must not become a new field, and its internal newline must not become a new record.
Replacing every comma with a semicolon would change the item name itself. Treating every newline as a record boundary would split the note. A separator conversion has to interpret the CSV structure before writing it again.
Check a round trip
Select all the source text and replace it with the semicolon preview above. Do not append the preview to the old input.
Now choose Semicolon (;) as the current separator and Comma (,) as the new separator. Convert again.
The displayed result returns to the original three records and their field values. The status again reports three rows. In the two data records, confirm:
• The item values are Brass, small and Steel.
• The first note still has its internal line break, semicolon and quoted word.
• The quantities remain 2 and 3 in the third field.
A round trip is a useful small check, not proof that every possible CSV dialect is supported. Equivalent CSV can use different optional quoting or line endings, so byte-for-byte comparison is not the same as comparing parsed data.
Before importing a real file
Keep the original file. Test a small, non-sensitive sample that includes the awkward fields your data actually uses, then check the destination application's delimiter setting and preview before accepting the import.
Separator conversion does not decide whether a field is a number, a date, text or a spreadsheet formula. It is not formula sanitization. The tool's documentation also says downloaded text uses UTF-8 and that it does not detect or convert legacy encodings.
RFC 4180's CSV format description explains the common comma-separated quoting convention. Semicolon-delimited files are a related dialect; the receiving program still needs to be configured for it.
Disclosure: This original tutorial was prepared by an AI assistant authorized by Hello Keyboard, the linked tool's operator. The comma-to-semicolon and reverse conversions above were checked in the browser on 7 November 2026 using invented text. This is a limited functional example, not an independent review or a security or performance audit.