Nikaia
Menu
v0.0.1343pre-alpha GitHub

agent: builder model: opus max_turns: 70 budget_usd: 18 —

Brief: #1015, nikaia fmt keeps a grammar rule line holding if(cond, "label") parseable

Score: a defect that blocks this lane (y2c run 2 stopped on it: the draft’s 9 rule lines over 100 columns with a labelled guard do not survive fmt, so the package gate’s fmt::every_nika_file is red). Runs first; y2c-properties.md (run 3) waits on it.

Issue (#1015, copied)

A long rule line with a guard if(cond, "label") is reflowed by nikaia fmt (crates/nikaia-std/src/tools/fmt.nika) into text that no longer parses. To build: (1) the formatter breaks guard arguments only where the grammar parser accepts a break (or keeps the guard on one line); (2) a case the fmt::every_nika_file set formats.

Probes (the architect ran both on today’s develop, target/fast/nikaia of 06:15)

fn f(a: i64, b: i64) -> bool sync {
	return a < b
}

grammar G {
	rule WS = "" -> { }
	entry rule top -> i64 = a:here abcdefghijklmnopqrstuvwxyzabcdefghijklmnopqrstuv:here if(f(1, 2), "a node its tag fits") eof -> { 1 }
}

fn main() {
	println("x")
}
  1. nikaia check passes; nikaia fmt refuses: “the formatted file does not parse: Expected ) here, but found ,” (line 9, column 24). The line is over 100 columns, reshape (fmt.nika:720) finds no top-level comma and no operator, and breaks the last bracket group that holds a comma, the guard’s if(...), through fmt_broken (:929), which puts each argument on its own line with a trailing , the grammar parser (winnow-grammar’s, not ours) does not take. The same line under 100 columns formats fine.
  2. The same rule broken by hand after a:here ...:here with the continuation at two tabs: check passes, fmt writes the continuation at one tab, the rule’s own level (a second layout defect: a continuation line belongs one level deeper than the line it continues, Part III nikaia fmt and how fmt breaks any other long line).

To build (two commits)

  1. No break that does not parse. A guard’s if(..) (and fail(..), any grammar call with arguments the grammar parser reads) inside a rule line is never broken by fmt_broken: treat it as sealed (fmt_sealed, :851, or the caller in reshape), so a long rule line breaks elsewhere or stays long (reshape returns [] when nothing breakable is left, as for any unbreakable line). Do not change winnow-grammar. Test tests/compiler/tests/fmt_keeps_a_grammar_guard_whole.nika + .stdout: probe 1 with the guard’s line over 100 columns, main calling G::top on an input it accepts and printing the result; fmt::every_nika_file formats it (it scans tests/ as it scans packages/, as fmt_keeps_a_nested_tuple_closure_whole.nika shows), and it is red on the unfixed binary (say so under Tests:).
  2. A rule’s continuation one level deeper. A rule line broken by hand (or by fmt, between sequence elements) keeps its continuation at the rule’s indentation + 1, and formatting again changes nothing (fixed point). Look at indent_of (:631) and the frames a grammar { } body pushes. Add probe 2’s shape to the step 1 test (a second rule, broken by hand) so the fixed point is checked.

fmt.nika lives under crates/nikaia-std/src/tools/: format it by hand (never nikaia fmt on it, builder-facts), then nikaia lower-std and commit its lowered twin crates/nikaia-std/src/tools/lowered/fmt.rs with it. Read docs/builder-facts.md (its first trap: the winnow-grammar rev of target/fast/nikaia).

Done when

Both probes format into a file that parses and is a fixed point; the new test passes; the gate of .claude/agents/builder.md (#852) and cargo test --profile fast -p nikaia --test it -- fmt:: green; then, as a check for the next brief, git fetch origin wip/y2c-properties && git show origin/wip/y2c-properties:packages/yaml/src/main.nika > $SCRATCH/main.nika and target/fast/nikaia fmt $SCRATCH/main.nika (after rebuilding target/fast/nikaia) formats and the result passes nikaia check (say so under Tests:; do not commit that file). Committed and pushed, step 2’s commit with Fixes #1015. Files you may change: fmt.nika, its lowered twin, the new test and its .stdout. Name under Changelog: what the change decided.

Rules (CLAUDE.md; this block is pasted into every brief, also a fix-forward)

Report (at most 15 lines, to exactly /home/user/Nikaia/docs/briefs/yaml/y15-1015-fmt-guard-label.md.report.md)

Fields of docs/architect.md, Briefs and reports: Result: done|noop|partial|blocked (the bare word alone on that line: Result: done, never a summary after it, which cuts the batch, b869a 2026-10-27; Owner: is a question for the owner or none, never a role), Files, Commit, Body (optional, one paragraph for the commit message), Fixes, Pushed, Tests, Gate (green|red), Deviation, Found, Needs, Changelog, Owner, Disk freed, and for a move Rs2nika: (the kept N of M lines of rs2nika --score, then the three costliest hand edits; none for a task that moves no Rust, none (redesign) for a brief with Shape: redesign and no Convert: line). none where nothing, and then the bare word: Deviation: none (a note) cuts the batch (batch.nika compares exactly); a note goes into the commit message. Found: is only for a defect outside your change (it cuts the batch); a process event such as another session resetting or rewriting your files is no Found:, it goes under Needs: (m3a 2026-10-28); a converter draft that came out wrong is no Found:, it is one of the costliest edits under Rs2nika:, which the converter lane reads (e12-4b 2026-10-27); how you solved your own step goes into the commit message. Deviation: is only a step of the brief built otherwise than written or not built (a choice the brief left to you, such as keep what X needs and say so, is no deviation: it goes into the commit message; nor is building two steps in one commit when every step is built, e12-3a 2026-10-27, nor a step another commit on develop already built, shown by your test failing with that commit’s change reverted, b889 2026-10-27: name the commit in Body:, nor a test’s expected .stdout/.stderr taken from what the program really prints where the brief’s sketch of it was short or wrong, b934 2026-10-29: say so in Body:, nor a lowering that follows the ADR governing it where the brief’s expected lowering contradicts that ADR, b933 2026-10-29: name the ADR in Body:, nor a lowering other than the brief’s where rustc cannot take the brief’s, b1029 2026-10-31: say why in Body:, nor re-keying an allow-list entry (KNOWN) whose hash changed for the same program and the same reason, b-fixforward-sweep2 2026-10-28: say so in Body:, nor an existing test’s expected .stdout/.stderr updated because your change rightly changes the message it prints, b1012b 2026-10-31: name the test in Body:); how you verified (e.g. a test not run against the unfixed compiler) goes under Tests:. A noop says why under Tests: (what was probed), not under Deviation:. Final answer, one line: done <sha>, report <path> (or BLOCKED: ... / PUSH REJECTED ...).

Attribution: Co-Authored-By: Claude Opus 5.5 noreply@anthropic.com Claude-Session: https://claude.ai/code/session_01UVxJDveRM5qKZKyCqq6Y24

This site is open source. Improve this page.