Thanks to visit codestin.com
Credit goes to github.com

Skip to content

docs: stale-test-binary trap + concurrency rules in CLAUDE.md; bump corpus #75

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
luisleo526 merged 1 commit into from
Jul 7, 2026
Merged

docs: stale-test-binary trap + concurrency rules in CLAUDE.md; bump corpus #75

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
24 changes: 24 additions & 0 deletions CLAUDE.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,8 +19,32 @@ ctest --test-dir build --output-on-failure
./scripts/run_corpus.sh
```

> **Stale-test-binary trap:** `cmake --build build --target pineforge` rebuilds
> ONLY the static lib. Test executables are separate targets that statically
> link it, and `ctest` does not rebuild anything — after a lib-only build,
> ctest runs STALE test binaries and can falsely pass. Always run a full
> `cmake --build build` (all targets) before `ctest` when runtime values
> changed. Before trusting any result, `build/lib/libpineforge.a` must be
> newer than every working-tree file in `src/`, `include/`, and
> `CMakeLists.txt` (uncommitted edits count) — else the conclusion is a
> stale-build artifact.

If `scripts/run_corpus.sh` reports any parity drift or failures, investigate the underlying cause. No regressions are allowed unless a resolved bug was previously masking a divergence.

## Concurrency

- `scripts/derive_corpus_feeds.py` (`ensure_derived()`) is a cheap no-op when
`corpus/data/derived/` is fresh, but a REBUILD is not concurrent-safe: it
writes through a fixed `*.csv.new` tmp then renames, so two processes
materializing simultaneously race and can corrupt the derived feeds. Run it
once to freshness BEFORE fanning out parallel consumers.
- Never run `scripts/run_corpus.sh` while any external harness that links
`build/lib/libpineforge.a` or reads `corpus/data/derived/` is running:
run_corpus.sh re-materializes the derived feeds and rebuilds the lib
(`--target corpus_strategies` links `pineforge`), clobbering both under the
concurrent run. External sweeps may run in parallel with each other over
disjoint work sets, but never overlapped with run_corpus.sh.

## Build Commands

- Configure CMake: `cmake -B build -S . -DPINEFORGE_BUILD_TESTS=ON -DPINEFORGE_BUILD_CORPUS_STRATEGIES=ON`
Expand Down
2 changes: 1 addition & 1 deletion corpus
Open in desktop
Submodule corpus updated 7 files
+16 −0 CLAUDE.md
+24 −20 README.md
+2 −2 draft-probes/README.md
+2 −2 validation/ta-hma-55-close-cross-01/README.md
+2 −2 validation/ta-macd-12-26-9-line-signal-cross-01/README.md
+3 −3 validation/ta-rsi14-cross-50-01/README.md
+2 −2 validation/ta-sma-152-close-cross-01/README.md
Loading