If a transaction description on your Cler bank statement is long enough to wrap across two or three lines, it still needs to land in a single row once converted. When a transaction description wraps across several lines on the statement document, the parser treats those lines as a single record instead of splitting them into phantom rows. It reads from the first date or amount column it recognises and keeps that row together until the next actual transaction begins, so a long narration that fills two or three lines does not create extra empty cells in the table.
If the running balance on your Cler bank statement does not add up row by row, that is exactly the kind of mismatch Convert·Into is built to catch rather than pass through silently. Every statement’s own printed closing balance is checked against the calculated running balance during extraction. If a row’s amount does not tally with the progression from opening to closing figure, that row is flagged for review rather than causing the whole file to be rejected, so you can confirm and correct a single mismatch without losing the rest of the data.