Issue #1052: fmt breaks a long grammar rule inside a repetition bound {2}

Open and answer this issue on GitHub.

ready · opened 2026-10-31 · updated 2026-10-31 · 0 comments

Found while writing grammar_hex_digit_is_one_hex_digit.nika (#1002).

grammar Pair {
	rule WS = "" -> { }
	entry rule top -> String = "#" r:hex_digit{2} g:hex_digit{2} b:hex_digit{2} eof -> { f"{r} {g} {b}" }
}

nikaia fmt turns the long line into

	entry rule top -> String = "#" r:hex_digit{
		2
	} g:hex_digit{2} b:hex_digit{2} eof -> { f"{r} {g} {b}" }

It still parses and runs the same, but the formatter reads the bound {2} as a block and breaks there. A bound is one token group and should never be broken; a long rule is broken between elements, or before ->.

To build

1. fmt keeps a repetition bound ({n}, {m,n}, {m,}) on one line, the same for digit, hex_digit and any element.
2. A test tests/compiler/tests/fmt_keeps_a_repetition_bound_whole.nika with the rule above, its .stdout the formatted text.

Back to Open Issues