IJHack / QtPass / 27473596278

57%

push

github

refactor: dedup executeBlocking + clarify isPathInStore guard (#1510) (#1520)

* refactor: dedup executeBlocking run loop + clarify isPathInStore guard (#1510)

Two small cleanups from the thermonuclear audit (#1508).

1. Executor::executeBlocking duplicated the start/stdin/wait/decode body
   between the input-taking overload and the env-aware overload. Extract
   a private Executor::runBlocking() helper and route both through it.
   No public signature change. The env-aware overload still parses its
   QStringList env into a QProcessEnvironment, then delegates.

   (The broader "collapse to two overloads taking QProcessEnvironment"
   from #1510 is deferred: changing the public env type ripples into
   Pass::env / Executor::environment() QStringList plumbing — not a small
   cleanup. Filed intent stays in #1510.)

2. Util::isPathInStore's ancestor-walk had a cdUp() guard the audit
   flagged as effectively dead. It is defensive (the filesystem root
   always exists, so the loop terminates before cdUp can fail). Document
   that invariant inline rather than removing the guard.

Behaviour unchanged. tst_util 138, tst_executor 32, tst_integration 20
all pass; doxygen clean.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

* fix: always close stdin in Executor::runBlocking

Previously stdin was only closed when input was non-empty (preserved
from the pre-refactor executeBlocking). A child that reads stdin until
EOF would hang on the empty-input path because the write channel stayed
open. Close it unconditionally — these are one-shot blocking runs that
never stream, so closing stdin is always correct.

tst_util 138 / tst_executor 32 / tst_integration 20 pass.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>

16 of 16 new or added lines in 1 file covered. (100.0%)

3731 of 6744 relevant lines covered (55.32%)

36.74 hits per line