Claude-Session: https://claude.ai/code/session_01Ji7xZMYjc4zankoc7hGBE1

- MODULE.bazel names the module progmasoft_catch3, the name it has in
  the registry, with bazel_compatibility >=8.0.0; //tools:version finds
  the version under the new name.
- .bcr/ holds the templates publish-to-bcr fills: metadata, source (the
  GitHub archive of the catch3-* tag) and the presubmit, which runs the
  test module on Debian, Ubuntu, macOS (x86-64 and arm64) and Windows
  with Bazel 8.x and 9.x.
- tests/Progmasoft/BazelModule is a module that depends on Catch3 as a
  user's does, through //:catch3_main and through //:catch3 with its
  own main. The root ignores it; Bazel Builds tests it with Bazel 8.x
  and 9.x on Ubuntu and Windows.
- Publish to BCR runs when a catch3-* release is published, or by hand
  for a tag, and opens the registry pull request from the fork of the
  maintainer, without attestations.
- The integration guide says how a module depends on progmasoft_catch3.

Claude-Session: https://claude.ai/code/session_01Ji7xZMYjc4zankoc7hGBE1

Every Progmasoft-authored file moves from MPL-2.0 WITH
AdditionRef-Progmasoft-Exception-2.0 to LGPL-3.0-or-later, from 0.5.0
on. Catch2-originated files keep BSL-1.0, and the vendored yyjson and
pugixml keep MIT.

- LICENSES/ holds the LGPL-3.0 text and the GPL-3.0 text it builds on,
  both as published by the FSF; the MPL-2.0 and Progmasoft Exception
  texts go, since no file uses them any more.
- Five files that came from Catch2 and were rewritten by Progmasoft
  (.clang-format, the issue and pull request templates, SECURITY.md)
  keep Catch2's notice next to Progmasoft's and say
  LGPL-3.0-or-later AND BSL-1.0.
- //tools:check expects the new header of Progmasoft files; the
  amalgamation notes say LGPL-3.0-or-later and the amalgamated files
  are regenerated; meson's project license is BSL-1.0 and
  LGPL-3.0-or-later.
- README gains a License section; LICENSES/README.md, contributing.md
  and progmasoft-catch3.md describe the new split and that releases
  before 0.5.0 keep the license they were published under.

Every commit to Progmasoft's files since the fork is the maintainer's.

Claude-Session: https://claude.ai/code/session_01Ji7xZMYjc4zankoc7hGBE1

- gcov is run with --branch-probabilities; without it the JSON has no
  branches, so GCC runs had none in coverage.json and coverage.lcov.
- PrepareCoverageDirectory also removes coverage.json, coverage.lcov and
  coverage.profdata of an earlier run, so a run that writes no reports
  leaves none to be read as its own.
- WriteCoverageReports returns notes and failures apart. A profile the
  tools cannot read or a report that cannot be written counts as a missing
  report and fails a run whose tests passed; no profiles or missing tools
  stay notes. The paths are announced only for the reports written.
- ProfiledAttempt takes the last attempt whose profile exists, so a final
  retry or repeat that crashed before writing one no longer hides the
  coverage of the attempt before it.
- The per-test llvm-profdata/llvm-cov and gcov runs use as many threads
  as the run had workers; results keep the order of the tests.

Tests: earlier reports removed and no failure without profiles; an
attempt without a profile gives way to an earlier one; the fixture gains a
branch taken one way, checked in the combined LCOV for both LLVM and gcov.
Amalgamation regenerated.

--coverage=DIR runs every attempt in a process of its own and keeps the
coverage of each test, for binaries built with LLVM source-based
coverage (Clang: -fprofile-instr-generate -fcoverage-mapping) or gcov
(GCC: --coverage).

How it works:

- The scheduler gives each attempt a fresh worker, as --isolate does,
  and sets LLVM_PROFILE_FILE and GCOV_PREFIX for it: the profile the
  process writes at exit is then the coverage of that one test, in
  DIR/profiles/t<index>-a<attempt>.profraw or a .gcov directory. Each
  runtime reads its own variable, so no symbol of either runtime is
  looked up: weak references would not pull them out of their static
  archives, and Mach-O does not allow undefined weak symbols.
- A process that runs other tests first (bisection) writes beside the
  profiles, never into one. The attempt records its profile, also in
  the JSON results.
- The profiles of an earlier run are removed when the run starts,
  because gcov adds to the .gcda files it finds.
- After the run, CoverageReport reads the last profile of every test:
  llvm-profdata and llvm-cov (PATH, xcrun on macOS, or
  CATCH3_LLVM_PROFDATA / CATCH3_LLVM_COV) export LCOV, gcov (or
  CATCH3_GCOV) writes JSON next to a copy of the .gcno. The result is
  DIR/coverage.json, the files, lines and branches each test reached,
  and DIR/coverage.lcov for the whole run. Without the tools the
  profiles stay and the run says so.
- --coverage with --in-process or --break is refused.

Tests: option parsing, LCOV reading and writing, gcov JSON, and an
integration test on CoverageFixture, a binary CMake builds with the
coverage flags of Clang or GCC: two tests reach different lines, each
map holds only its own, the LCOV report holds both, and a stale profile
of an earlier run is gone. Locally all 84 CTest tests pass with Clang,
and the coverage tests pass with GCC (built with
CATCH3_ENABLE_WERROR=OFF: Catch3 does not build with GCC's warnings as
errors, before and after this change).

CI: the Linux jobs install LLVM's coverage tools, a new job checks the
gcov path with GCC, and CATCH3_COVERAGE_REQUIRE_MAP turns a missing map
into a failure on Linux and macOS instead of a skip. MSVC has no
source-based coverage, so its builds skip those tests; the roadmap
lists step 4 as in progress until Windows with clang-cl is tested.

Workers and the other processes the runner starts copied its whole
environment, tokens and passwords included. ChildProcess::Start now
leaves out every variable whose name contains, in any case, TOKEN,
SECRET, PASSWORD, PASSWD, CREDENTIAL, API_KEY, APIKEY, PRIVATE_KEY or
ACCESS_KEY (IsSecretVariable), unless ChildOptions::keptEnvironment
names it; names compare without case on Windows.

--pass-env=NAME, repeatable, and "pass-env" in catch3.json keep a
variable for the tests. The runner itself still starts with the whole
environment, so code that runs before the first worker sees it;
docs/progmasoft-runner.md says so.

A runner test starts the shell's environment listing with a secret and
a plain variable set, sees only the plain one, and sees the secret once
it is kept.

A report no longer throws, so a noexcept mock survives an aborting run.
A call that has nothing to return still leaves the test case, which ends
the process in a noexcept function; the documentation says so.

RunContext clears the result capture when it goes, so a mock of static
lifetime finds none and writes to standard error.

Decision owns its source line, since Verify can free the expectation
while a call is on its way. Moving a result out takes the lock. Vector
matchers refer to their vector, so they are rejected at compile time
like composed ones. A null text expectation accepts only null, and
integers compare by value, so 256 no longer accepts an unsigned char.

A static mutex needs an exit-time destructor, which Clang's warnings
refuse, and a mock of static lifetime can outlive it. The test seam is
held the same way, and a test covers a mock method with nothing after
its parameters, which needs an empty variadic argument to stay valid
under -Wpedantic.

Step 3 of the roadmap. MockFunction checks the calls of one function
against its expectations; CATCH3_MOCK_METHOD declares one for a virtual
member, and CATCH3_MOCK_FUNCTION and ScopedSeam serve free functions
behind a seam.

Expectations take values, Catch matchers, Any or Where; count calls with
Times, Between, AtLeast, AtMost, AnyTimes and Never; answer with Returns,
Does or Throws; and order calls across mocks with Sequence. A failure
goes through AssertionHandler at the line of the expectation, so every
report of the runner shows it. Met expectations count as passed
assertions. A test case that is already being left does not get unmet
expectations on top of its own failure.

Composed matchers refer to temporaries, so they are rejected at compile
time rather than left dangling. A call that fails returns a
value-initialized result, or leaves the test case when no such result
exists.

The failure paths are tested through the runner fixture, reading the
messages from its JSON report.

A review of the import found ways for a failure to disappear or for the
import to reach too far.

- CTRF's retryAttempts ends before the final attempt, which was dropped when
  its outcome matched the one before it.
- TAP names tests only within a stream, so a failure in one file and a pass
  under the same name in another merged into a flaky pass. The path given to
  --import is now the class of a stream's tests.
- A failure followed by a skip with the same identifier read as skipped.
- Directives are whole words: TODOING is not TODO. SKIPPED and SKIPPING stay
  directives, as harnesses treat them.
- Verdicts that alternate count as flaky only under an explicit
  --import-revision. HEAD says nothing about a dirty checkout or a report from
  another one, and two such events could quarantine a real failure.
- --import-name no longer redirects the history of a normal test run.
- Importing does not depend on the tests registered in the binary.