ci: cache generated test fixtures keyed on toolchain and inputs - #436

The fixture-cache step merged in #436 landed after this branch was cut
and tripped the new zizmor gate on the PR's merge commit.

Co-Authored-By: Claude Fable 5 <[email protected]>

* Replace SonarCloud with an in-repo zizmor gate

SonarCloud had degraded to an untriaged dashboard: a red quality gate
nobody was watching, a violations badge advertising it, and no PR
integration. Its only findings that SwiftLint/SwiftFormat/ShellCheck
could not have made were about the workflows themselves, so cover that
gap in-repo instead:

- Add a zizmor job to lint.yml, gated at low severity and above so it
  starts green; informational findings (tag-derived template expansions
  in release.yml) are deliberately below the floor.
- Pin the official actions (checkout, upload-artifact, download-artifact)
  to commit SHAs, which was SonarCloud's one open gate-worthy finding
  and is required by zizmor's blanket pinning policy.
- Drop the SonarCloud badge from the README.

The bump_version checkout keeps its credentials on purpose, so it
carries an inline artipacked ignore rather than a fix.

Co-Authored-By: Claude Fable 5 <[email protected]>

* Pin actions/cache, newly arrived from main

The fixture-cache step merged in #436 landed after this branch was cut
and tripped the new zizmor gate on the PR's merge commit.

Co-Authored-By: Claude Fable 5 <[email protected]>

---------

Co-authored-by: Claude Fable 5 <[email protected]>

A scheduled workflow (Mon/Thu + workflow_dispatch) that exercises the
newest Xcode on the runners end to end: checks that xcresulttool still
advertises legacy command support (the early-warning signal for the #391
migration), builds the package, regenerates fixtures from scratch via
prepareTestResults.sh, runs the test suite, and asserts xchtmlreport
exits clean against a bundle produced by that exact toolchain. Any fault
files or updates a single open `drift`-labelled issue, so a break cannot
rot silently in the Actions tab.

Fixtures are deliberately regenerated (never cache-restored) because the
job's purpose is to exercise the current toolchain — but a successful
run saves them under the #436 cache key, pre-warming PR jobs after a
toolchain bump.

Also folds in the #436 fast-follow: the fixture cache key derived its
runtime from `simctl list runtimes` (newest installed) while
prepareTestResults.sh selects the newest runtime that offers an iPhone.
The selection now lives in scripts/select_simulator.py and both consume
it, so the key can never disagree with the runtime the script boots.

Closes #392

Co-Authored-By: Claude Fable 5 <[email protected]>

* ci: add scheduled toolchain drift detector (#392)

A scheduled workflow (Mon/Thu + workflow_dispatch) that exercises the
newest Xcode on the runners end to end: checks that xcresulttool still
advertises legacy command support (the early-warning signal for the #391
migration), builds the package, regenerates fixtures from scratch via
prepareTestResults.sh, runs the test suite, and asserts xchtmlreport
exits clean against a bundle produced by that exact toolchain. Any fault
files or updates a single open `drift`-labelled issue, so a break cannot
rot silently in the Actions tab.

Fixtures are deliberately regenerated (never cache-restored) because the
job's purpose is to exercise the current toolchain — but a successful
run saves them under the #436 cache key, pre-warming PR jobs after a
toolchain bump.

Also folds in the #436 fast-follow: the fixture cache key derived its
runtime from `simctl list runtimes` (newest installed) while
prepareTestResults.sh selects the newest runtime that offers an iPhone.
The selection now lives in scripts/select_simulator.py and both consume
it, so the key can never disagree with the runtime the script boots.

Closes #392

Co-Authored-By: Claude Fable 5 <[email protected]>

* ci: keep the legacy alarm reachable when xcresulttool itself vanishes

Review follow-up (CodeRabbit): the toolchain recording step invoked
`xcrun xcresulttool version` bare, so the tool disappearing entirely
would fail recording, skip the legacy probe, and file a generic setup
failure instead of the one alarm the workflow exists to raise. Recording
now tolerates the failure and the probe swallows the command error so
the missing marker fires the explicit ::error either way.

Co-Authored-By: Claude Fable 5 <[email protected]>

* ci: compute the fixture cache key before generation via a shared action

Review of #442 found the drift job's pre-warm dead on arrival: its save
key rendered hashFiles('XCTestHTMLReportSampleApp/**') at save time,
after prepareTestResults.sh had left .derivedData inside the hashed
tree (actions/glob matches dot-dirs), so the saved key could never
equal test.yml's clean-tree restore key — ~95 MB of unreachable cache
churned per run.

Both workflows now derive the key from one composite action
(.github/actions/fixture-cache-key) that runs before generation. The
hash inputs and key format are unchanged, so existing cache entries
still hit; sharing the action also removes the duplicated key logic
the two workflows could silently de-sync on.

Also from the review:
- run the legacy probe under !cancelled() so a hard-failed record step
  (xcodebuild/swift dying — the alarm's most important case) cannot
  skip it, and have the report call out a probe that somehow still
  did not run
- make the record step's xcresulttool tolerance survive pipefail
  (|| true inside the pipeline) instead of relying on the default
  shell leaving pipefail off

Co-Authored-By: Claude Fable 5 <[email protected]>

---------

Co-authored-by: Claude Fable 5 <[email protected]>

…t + forced-modern CI leg (#391, Tasks 12–13) (#450)

* Parity rulings from the first differential run; fixes the #449 export race

Running both readers over freshly generated fixtures surfaced divergences
the spec's tables did not anticipate. Each lands here as a rule, recorded
in the spec's "Task 12 execution rules" section; none adds an allow-list
entry.

- Content-addressed attachment exports: both backends name every exported
  payload <payloadId>.<ext> — the one attachment identifier the two
  formats share. Xcode 26.2 gives every auto screen recording in a session
  one shared display name, so name-keyed exports collapsed distinct
  payloads onto one path and raced concurrent copies (#449, reproduced
  with a live EEXIST trace and a 1-in-12 test flake). The export is now
  idempotent: an existing destination is the payload, never removed.
- Symbol-annotation rows (no startTime) and attachment-shadow rows
  (leaf, non-failure, startTime == sibling attachment timestamp) are
  dropped by the modern reader; both are 26.2 bookkeeping the legacy tree
  never had. The shadow join's guards are load-bearing: a genuine failure
  row shares the attachment's millisecond on testWithSpecialChars().
- Swift Testing names come from the identifier's function form on both
  backends; the @test display name is a field only one backend can fill.
- Legacy merges parameterized argument executions (duplicate siblings
  with no repetitionPolicySummary) into one iteration; true retries keep
  their numbers.
- Expected failures are non-events on both backends: messages claim and
  remove their exact-title activity rows instead of joining.
- Skip notices were a legacy reader gap: the reason was always on
  skipNoticeSummary; it now renders the same appended row modern emits.
- Failure-row placement goes through one shared total-ordered interleave
  (ParsedActivity.interleavingFailureRows) fed by both readers; the
  modern reader hoists failure tips (isFailure now means "is the
  assertion row" — tip of the flagged chain). Re-nesting legacy rows by
  time window was tested and rejected: windows collide at millisecond
  granularity and misplace rows.
- Run logs export as <run-identifier-digest>.log on both backends instead
  of backend-internal reference names.

Refs #391. Fixes #449.

Co-Authored-By: Claude Fable 5 <[email protected]>

* test: the differential — legacy/modern parity held to a declared allow-list

Renders each fixture through both readers, forced explicitly, and asserts
(Task 12): per-run counts, the identifier→status map, and exported
attachment filenames AND bytes match exactly; after normalizing identifier
digests (#430's [0-9a-f]{32}, .linking renders only) and masking exactly
the four declared losses, the two renders are byte-identical. Skips loudly
when the toolchain has no legacy commands, and proves the forced backend
actually resolved rather than assuming it.

Anti-rot runs in both directions: a diff outside the allow-list fails the
build, and an entry that masks nothing real also fails — "fires" means
omitting the rule (others still applied) leaves the renders unequal, so an
entry cannot rot silently once Apple fills the gap. All four entries fire:
durations on all three fixtures, wrapperGroups on all three,
failureTitlePrefix and attachmentDisplayNames on TestResults and
RetryResults.

The masker diverged from the plan's snippets in four evidence-forced ways,
recorded in the plan's Task 12 execution note: wrapperGroups is a
structural SwiftSoup unwrap (line-filtering left the wrapper's div
skeleton behind), the display-name anchor spans the icon line the plan's
regex tripped on, line joins are canonicalized before comparing (template
concatenation breaks lines differently across nesting depths), and
XCTest-case duration coverage lost to the over-broad durations mask is
restored by a model-level assertion.

Refs #391.

Co-Authored-By: Claude Fable 5 <[email protected]>

* ci: forced-modern test leg via XCHR_RESULT_READER (Task 13)

One matrix leg runs the whole suite with the modern reader forced, so the
path that becomes the only path once Apple removes the legacy commands is
exercised end to end on every PR — not only through the differential.

The env var is the CI override, not a new control surface: it defaults
both Summary.init's backend and the CLI's --result-reader (so CLI-driven
tests pick it up through the spawned binary), the flag still wins when
passed, and an unrecognised value degrades to auto. Verified locally:
92 tests green under both XCHR_RESULT_READER=modern and =auto. Fixture
cache (#436) is shared across legs unchanged; no new action steps.

Refs #391.

Co-Authored-By: Claude Fable 5 <[email protected]>

* docs: record the Task 12 rulings as rules, and the 4.0 release-note items

Spec: new "Task 12 execution rules (2026-08-12)" section — content-addressed
exports (fixes #449), symbol-annotation and attachment-shadow drops, Swift
Testing names from the identifier, parameterized-execution merge, expected
failures as non-events, skip-notice fix, the shared failure-row interleave
with the reversible hoist, and run-log naming — each with the evidence that
forced it. The attachment-filename table row is superseded, the Tree-shape
display-name paragraph becomes a model rule, and attachmentDisplayNames'
Exercised-by cell gains RetryResults (measured).

Plan: Task 12 execution note (where the shipped harness supersedes the
snippets, and why); Task 15 gains a release-notes checklist so the 4.0
output changes accepted here cannot be missed by the docs task.

Refs #391.

Co-Authored-By: Claude Fable 5 <[email protected]>

---------

Co-authored-by: Claude Fable 5 <[email protected]>