fix(guard): spec-conformance fixtures are bytes — the zero-width key rendered only under a UTF-8 locale - #3109
Open
noahgift wants to merge 2 commits into
Open
fix(guard): spec-conformance fixtures are bytes — the zero-width key rendered only under a UTF-8 locale#3109noahgift wants to merge 2 commits into
noahgift wants to merge 2 commits into
Conversation
…rendered only under a UTF-8 locale The ledger_respend_zero_width_key fixture spelled its zero-width space as a \u200b escape and rendered it with printf '%b'. bash expands \u only when the locale is UTF-8; under a POSIX locale it keeps the six characters as literal text, so the key never collides with "lambda" and the row reads CLEAN. intel's runners run en_US.UTF-8, yoga's runner image ran POSIX, and main's guard-tree went RED on yoga-build3 (run 34529896596, job 103047791546: "BROKE ledger_respend_zero_width_key expected L1 got CLEAN"). The fixture now spells the bytes (\xe2\x80\x8b), which printf '%b' renders identically under any locale, and selftest() pins LC_ALL=C so a \u fixture fails on every box instead of only on the odd runner. Measured on a75e154's tree: before, 40/40 under C.UTF-8 and 39/40 with no locale; after, 40/40 under both. Mutation: restoring the \u200b escape goes RED (39/40) under a C.UTF-8 ambient locale once the selftest pins C. Pmat-Ticket: PMAT-1098 Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
noahgift
enabled auto-merge
September 10, 2026 22:59
|
§13.11 rung 1 — quorum shadow verdict Shadow mode: this records a verdict and merges nothing. A refusal |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What broke
Main's
guard-treewent RED on yoga-build3 (run 34529896596, job 103047791546):BROKE ledger_respend_zero_width_key expected L1 got CLEAN. The same tree passes 40/40 on intel.Why
The fixture spelled its zero-width space as
\u200band rendered it withprintf '%b'. bash expands\uonly under a UTF-8 locale; under POSIX it keeps the literal text, so the key never matches
lambdaand therespend row reads CLEAN. intel runs en_US.UTF-8; yoga's runner image ran POSIX (fixed on the runner side
since:
LANG=C.UTF-8in the yoga and gx10 runner env).Fix
\xe2\x80\x8b), identical under any locale.selftest()pinsLC_ALL=C, so a\ufixture fails on every box, not only on the odd runner.Evidence (a75e154's tree, run locally)
\u200brestored, selftest pins Cbashrs finding count unchanged. Pre-push:
cargo fmt --all -- --check,cargo test -p aprender-contracts --lib,cargo deny check advisoriesall green.Milestone 0.67.0: runner parity for the three-host build pool (intel + yoga + gx10).
🤖 Generated with Claude Code