Encoding first. A byte-order mark identifies UTF-8, UTF-16, or UTF-32. Without one, LedgerPair reads strict UTF-8 and falls back to Windows-1252, the code page older Windows accounting software writes, so names like Müller and Café come through as written. A file that is not text at all, such as an image renamed to .csv, is refused with a message that names it.
Delimiters and quoting. Commas, semicolons, tabs, and pipes are detected from the file. Quoted fields follow RFC 4180, so an address with a comma or a line break inside quotes stays in one cell. Blank lines are skipped, and a record with too many or too few fields is padded or trimmed with a warning that names the file and the row, recorded in the Notes on the Summary sheet.
Row numbers you can find. The header is row 1 and each record carries the row number a spreadsheet shows for it, so you can go straight to the same record when you open the file in Excel. The table shows the row in file A and the row in file B for every difference.
Keys. Tick one or several key columns in each file. The names can differ, such as Order in one export and Order No in the other, as long as the count matches. Keys are compared without regard to case, thousands separators and a trailing .0 are ignored, and leading zeros are kept, so 00123 and 123 stay distinct identifiers.
Values. Every column with the same heading in both files is compared cell by cell, and headings match regardless of case and extra spaces. Leading, trailing, and repeated spaces are ignored, amounts compare as exact decimals whatever their separators or currency symbol, and dates compare as calendar dates whatever their notation. A value such as 007 is kept as text, so it stays distinct from 7.
What you see. The table lists each difference with its kind, key, row in A, row in B, and a detail such as 1 cell differs: Price. Select a row and the pane below shows every compared column from both files, with the changed cell highlighted. The filter narrows the list to changed rows, rows only in file A or file B, or duplicate keys.
What you keep. Export evidence workbook saves a new .xlsx with Summary, Additions, Deletions, Changes, and Duplicates sheets. Each changed cell gets its own row on the Changes sheet, with the raw value from each file and the normalized form that was compared, so anyone can see why a cell was flagged.