Repository navigation
test(analog): 151 asserts on constant table elements now count, 13 specs, seals re-saved (Closes #12166) - #12191
Merged
Merged
Conversation
…ecs, counted total 86128 to 86279 (Refs #12166) An assert on a constant table element is folded by the compiler and tri test counts it as no assert at all (found in sar_kick, fixed there by #12153). Each such assert in the 13 other analog specs now reads the element into a var and asserts on the var, with the expected value unchanged. Measured with tri test per spec: the counted total of every spec rose by exactly the number of asserts rewritten (151 in all), every spec passes, and vacuous passes stay 0. Two negative controls per spec on scratch copies: one wrong table element turns the holding test red, one wrong expected value turns exactly the one test that holds it red. The note lists what is not claimed: seals are not re-saved here, strict coverage and the mutation census are not re-run, and the grep that found the candidates is a superset that misses lower-case and multi-line forms. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
…aster-built t27c (Refs #12166) Seals saved with a t27c built from master (CARGO_TARGET_DIR target-m2); strict coverage of specs/analog reads gaps 0, 130/130. The first save used a t27c older than master's gen-zig and gen-rust commits and moved 20 seals; this one moves exactly the 13 spec seals. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
…run (Refs #12166) Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Contributor
This was referenced Oct 25, 2026
This was referenced Oct 25, 2026
Closed
docs(analog): mutation census of the remaining analog specs, 8 of 17 written up (Refs #12116)
#12208
Merged
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
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
An assert on a module-constant table element,
assert_eq(TABLE[i], K), is folded at compile time andtri testdoes not count it. This PR reads each such element into avarfirst, so the assert runs and is counted.specs/analog: 86128 to 86279 (+151).docs/analog/assert-count-audit-2026-10-25.md.t27cbuilt from master. Strict coverage ofspecs/analog/*.t27readsgaps 0,coverage 130/130, tests 26/26, vacuous 0.Anomaly found on the way
The first seal job used a
t27colder than master's gen-zig and gen-rust commits: it moved 20 seals and strict coverage showedsar_logic(untouched here) as seal NO-MATCH. A master-equivalentt27cwas built in a T27 OS pane (CARGO_TARGET_DIR=/Users/playra/t27-build/target-m2), the seals were restored and re-saved, and this time exactly the 13 spec seals moved and every importer check passed. The lesson is in PR #12189.Not done
sar_cdacandsar_comparator_psp; their accepted-file lines still name the same code, the edits sit below them.Closes #12166
🤖 Generated with Claude Code