release: v0.1.2 — the gate-hardening batch + pre-release sweep - #968
Merged
Conversation
Rolls [Unreleased] (the four merged community PRs: #852 limitations-sync, #702 arm64 advisory lane, #956+#955 gate-honesty pair) into [0.1.2], plus what the pre-release sweep found: - fix(transform): an invalid string escape no longer crashes vera check with a raw lark.VisitError traceback (#966) — transform() unwraps a VisitError carrying a VeraError; E009 diagnostics gain the full instruction format (rationale/fix/spec_ref for the escape and interpolation classes), retiring a false grammar-prevents waiver. Test-first: text + JSON envelope tests, RED-confirmed pre-fix. - ci: two workarounds retired on fired triggers (#537) — pygments 2.20.0 shipped the CVE-2026-4539 fix (--ignore-vuln removed), and setup-python's toolcache carries pip 26.1.2 natively (--upgrade pip removed); both KNOWN_ISSUES rows retire with them. - docs: TESTING.md contract-verification block regenerated from live vera verify --json (283/378 obligations, 74.9% static - was three campaigns stale at 256/280); the summary off-by-one it surfaced is filed as #967 (KNOWN_ISSUES limitation + ROADMAP Stage 19 row); scheduled limitations-sync workflow documented; gate rows scoped to the #955 semantics; spec-block counts to 189; ~93,500 test-code lines; spec 0.5.1 acknowledges W001/W002; HISTORY's v0.1.0 column restored to its true snapshot (6,779) with the growth chart in lockstep; README/FAQ counts to 6,848; three test-code comments citing since-fixed issues (#918, #706, #635) retired. Release machinery: versions to 0.1.2 (10-file surface), HISTORY gains the combined 'Stage 19 and Stage 20' dual-threaded section + index row, 196-day... 96-day count recomputed from git, 200 tagged releases. Closes #966 Closes #537 Co-Authored-By: Claude <noreply@anthropic.invalid>
📝 WalkthroughWalkthroughThis PR fixes invalid string-escape crashes by unwrapping ChangesE009 Diagnostic and VisitError Unwrapping
v0.1.2 Version Bump and Release Documentation
Dependency-Audit Workaround Removal
Estimated code review effort: 3 (Moderate) | ~25 minutes Sequence Diagram(s)sequenceDiagram
participant cmd_check
participant transform
participant LarkTransformer
participant decode_string_escapes
cmd_check->>transform: transform(tree)
transform->>LarkTransformer: visit tree
LarkTransformer->>decode_string_escapes: process STRING_LIT token
decode_string_escapes->>decode_string_escapes: raise _escape_error (E009)
LarkTransformer->>transform: raise VisitError(orig_exc=VeraError)
transform->>transform: unwrap VisitError, re-raise VeraError
transform->>cmd_check: propagate VeraError with rationale/fix/spec_ref
Possibly related PRs
Suggested labels: 🚥 Pre-merge checks | ✅ 7 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (7 passed)
✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
Comment |
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #968 +/- ##
==========================================
+ Coverage 93.23% 93.30% +0.06%
==========================================
Files 96 96
Lines 31448 31470 +22
Branches 456 456
==========================================
+ Hits 29321 29362 +41
+ Misses 2114 2095 -19
Partials 13 13
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
|
Caution Failed to replace (edit) comment. This is likely due to insufficient permissions or the comment being deleted. Error details |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@vera/transform.py`:
- Around line 395-408: _parse_interp_expr() is still bypassing the new
VisitError unwrapping path, so E009 inside string interpolation can be
double-wrapped and escape correctly handled. Update the interpolation parsing
flow in VeraTransformer.transform() / _parse_interp_expr() so the same
VisitError boundary is used there, or explicitly flatten any nested VisitError
chain before re-raising. Make sure the fix preserves the existing
invalid-expression and statements-not-allowed errors while preventing wrapped
interpolation errors from slipping through.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: ASSERTIVE
Plan: Pro Plus
Run ID: e9ffb66f-97bc-44b0-85ca-4976263b9243
⛔ Files ignored due to path filters (7)
assets/diagrams/history-growth.svgis excluded by!**/*.svgdocs/index.htmlis excluded by!docs/**docs/index.mdis excluded by!docs/**docs/llms-full.txtis excluded by!docs/**docs/llms.txtis excluded by!docs/**tests/conformance/ch09_show_hash_composites.verais excluded by!**/*.verauv.lockis excluded by!**/*.lock,!uv.lock
📒 Files selected for processing (16)
.github/workflows/ci.ymlCHANGELOG.mdFAQ.mdHISTORY.mdKNOWN_ISSUES.mdREADME.mdROADMAP.mdTESTING.mdpyproject.tomlscripts/check_diagnostic_fields.pyspec/00-introduction.mdtests/test_cli.pytests/test_codegen_gc_rooting.pytests/test_codegen_interpolation.pyvera/__init__.pyvera/transform.py
🔗 Linked repositories identified
CodeRabbit considers these linked repositories for cross-repo context during reviews:
aallan/vera-bench(manual)
… ongoing-thread (PR #968 review)
… (PR #968 review) A nested transformer (_parse_interp_expr's inner VeraTransformer runs inside the outer transformer's STRING_LIT callback) can double-wrap a VeraError; the single-level unwrap would re-raise the wrapper. Not user-reachable today - interpolation segments cannot contain nested string literals (the outer lexer closes at the first quote, verified by probe), and the inner transformer's other token callbacks cannot raise on grammar-valid input - but _unwrap_visit_error now walks the whole chain (unit-tested with a hand-built double wrap, RED-first), and the user-reachable interpolation-statement E009 path gains its own envelope pin. Counts follow (+3 tests).
…roject's motion, not a headcount
aallan
changed the title
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
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.
The v0.1.2 release: rolls the four merged community PRs into
[0.1.2], plus everything the six-auditor pre-release sweep surfaced — including one genuine crash bug, fixed test-first in this PR.The batch being released
ubuntu-24.04-armto CI matrix to actually test Linux aarch64 #702 — advisoryubuntu-24.04-armCI lane (wheel gate + platform policy in lockstep)# diag-fields-exemptopt-out is honoured by only one of three passes #955 — the diagnostic-fields gate-honesty pair (reachability election with generic rebind counting; opt-out scoped to unresolvable fields)Four PRs — #960, #962, #964, #965 — closing three Stage 20 rows plus one ongoing-thread item, all by @chethanuk and each hardened in review.
What the sweep found (23 confirmed findings, 2-skeptic verified)
Fixed in this PR:
vera checkcrashed with a rawlark.VisitErrortraceback on an invalid string escape (Closes vera check crashes with a raw lark.VisitError traceback on an invalid string escape (E009) #966). The E009TransformErroris raised inside a token callback, which lark wraps — soexcept VeraErrornever fired and--jsonmode emitted zero bytes for CI to gate on.transform()now unwraps and re-raises; E009 diagnostics gain the full instruction format (rationale/fix/spec_ref for both escape and interpolation classes), retiring a# diag-fields-exemptwaiver whose "the grammar prevents this" premise was false. RED-confirmed text + JSON tests.--ignore-vulnremoved — left in place it would mask a real regression), and setup-python's toolcache now carries pip 26.1.2 natively (--upgrade pipremoved). Both KNOWN_ISSUES rows retire per the bridges-not-permanent-exceptions contract.vera verify --json, which also surfaced a summary off-by-one on three examples — filed as vera verify --json summary: tier1_verified + tier3_runtime = total + 1 on three examples #967 (KNOWN_ISSUES limitation + ROADMAP Stage 19 row: derive the summary from the reified obligations list).W001/W002; the scheduled workflow documented in TESTING.md; gate rows scoped to the Diagnostic-fields gate: the# diag-fields-exemptopt-out is honoured by only one of three passes #955 semantics; three test-code comments citing since-fixed issues (Verifier crashes with uncaught Z3 Sort mismatch on a nested-Option-returning function (check-green, run-OK, verify-crash) #918, Move Map/Set host stores from Python mirror to bucket-as-truth #706, Parameterised-alias substitution missing in _type_expr_to_wasm_type (compilability check) #635) retired.Release machinery
Versions to 0.1.2 across the 10-file surface;
[Unreleased]emptied into[0.1.2]with compare links; HISTORY opens the combined "Stage 19 and Stage 20: The dual-threaded start" section (maintainer decision: the external thread runs the single-source sprint while verification completeness follows) with its index row; 200 tagged releases / 96 active days, both computed from git.Closes #966
Closes #537
🤖 Generated with Claude Code
Summary by CodeRabbit