If a transaction description on your Swedbank statement is long enough to wrap across two or three lines, it still needs to land in a single row once converted. Upload a statement document and the engine starts work immediately. When a transaction description wraps across multiple lines, the system reads each continuation line as part of the same record rather than treating it as a separate row. It does this by detecting the repeating structure of the transaction table: a new entry always begins with a date, an amount, or a sequence number, while anything that follows without that signature belongs to the previous entry. The result is one clean row per transaction, with the full description intact, instead of fragmented text that would need manual re-stitching in a spreadsheet.
However your Swedbank statement lays out its columns, that layout is only useful to you once it is mapped correctly into the export. The engine identifies every column in the statement table and maps each one into a standard field in the export layout. Date, description, debit, credit and running balance all land in predictable columns every time. You can adjust the mapping before export: drag a column to a different position, rename a header, or exclude a field you do not need. The layout you define is saved, so the next statement from the same source uses the same structure without any setup.
If the Swedbank statement you have is a scanned image or a phone photo instead of an exported file, that alone does not stop Convert·Into from reading the transaction table off it. Scanned and photographed statements are read with OCR, and the same column mapping and export logic applies. The running-balance check is especially useful here: any character the scan misreads that breaks the arithmetic is flagged by row, so you can correct it before the data leaves the page. No file is rejected wholesale for a single misread digit.