Bank statement PDF to Excel

Why most PDF to Excel conversions come out broken

Copy the text out of a statement PDF and you get a stream of words in reading order. The moment a row has an empty cell, or an amount sits further right on one line than another, that stream loses the relationship between the date, the description and the amount. You end up with a column of dates followed by a column of amounts, unrelated to each other.

• Empty cells collapse, so a row with no description shifts every later value one column left.
• Amounts are right-aligned for print, so their starting position varies with the number of digits.
• Text is emitted in the order the PDF writer chose, which is not always left to right.

What this converter does instead

Every glyph carries a position. Text sharing a vertical band is grouped into a visual line, then sorted left to right within that line. That reconstructs the printed table before any parsing happens — so an empty cell stays empty and a right-aligned amount stays in its own column.

When the PDF has no text at all

Scanned statements, and PDFs printed from an image, carry no glyphs to position — only a picture of a page. Those need to go through image input instead: upload the PNG or JPG of the transaction page and the text is read visually. The converter detects this case and tells you rather than returning an empty sheet.