Splitter.SplitTextByPositions
Returns a splitter that cuts text at fixed character positions, for fixed-width files and layouts without delimiters.
Splitter.SplitTextByPositions(positions)Returns: A function that you pass to a split step. It returns a list of pieces.
What it does
Some reports have no delimiter at all. Each field occupies a set number of characters, which is called fixed width. Splitter.SplitTextByPositions cuts the text at the character positions you list. Positions count from zero, so position 0 is the first character.
Syntax and arguments
positions: a list of starting positions, from zero, such as{0, 8, 12}. Each position starts a new piece, and the last piece runs to the end of the text.
Example
A legacy ledger export writes each line as a date of eight characters, an account code of four characters and an amount. The date is in characters 0 to 7, the account in 8 to 11, and the rest is the amount.
let
Line = "20260331100012500",
Pieces = Splitter.SplitTextByPositions({0, 8, 12})(Line),
Date = Pieces{0},
Account = Pieces{1},
Amount = Pieces{2}
in
Amount
Pieces is {"20260331", "1000", "12500"}, so Amount is the text "12500". The conversion to a number and a decimal point would be a separate step.
Example: a second, different use
Applying the same positions to a table column with Table.SplitColumn, so each fixed-width field becomes its own column.
Things to watch
- Count characters, not visible width. A stray space or a special character shifts every later field.
- Check the last line of the file. Totals, footers and page breaks often have a different layout.
- Trim each piece after splitting, because padding spaces are common in fixed-width files.
- For a whole file, read the lines first with Lines.FromBinary, then split each line.
In business terms
Mainframe reports, older general ledger exports and government or bank formats are often fixed width. The layout is usually documented in the system guide, and a reviewer can check a few lines by hand against the positions.