Lesson 04 · Line and column positions

Line and column positions

A byte offset is precise but unfriendly - people read errors as "line 2, column 3". Today you track line and column alongside the offset so every token knows its human-readable position.

The goal

Stamp each token with a 1-based line and column, resetting the column after every newline.

Start here - the target

TO DO

Scenario: Positions across a newline

Giventhe input "[\n ]" (a bracket, a newline, two spaces, a bracket)

Whenit is scanned

Thenthe LBracket is at Line 1, Column 1, and the RBracket is at Line 2, Column 3

Andline and column both start at 1, a line feed increments Line and resets Column to 1, and a tab advances Column by 1

Background

A byte offset is exact but nobody debugs by counting bytes. Editors and error messages speak in line and column, so the scanner tracks those too. Both start at 1. As the scanner walks forward it counts columns, and every time it passes a line feed (\n) it bumps the line and snaps the column back to 1.

Keep the rule mechanical: advancing over one byte either moves the column right or, for a newline, drops to the next line. A tab counts as a single column here - not a jump to the next tab stop - which keeps positions predictable regardless of how an editor renders tabs. Record the line and column at the point each token starts, the same moment you already record its offset, so a token carries all three coordinates at once.

Make it work

// add Line and Col to Token; track them as you advance
// start line=1, col=1
// on consuming a byte: if it is '\n' { line++; col = 1 } else { col++ }
// record line and col at the moment a token starts (after skipping ws)

CheckpointDONE

Every token reports a human-readable line and column. Commit and stop here.