IJHack / QtPass / 35987557902

93%

Build Type

push

github

Committed by web-flow

Commit Message

Fewer branches where CodeFactor counts them (#1916)

* One table says what a subprocess is called and whether its output is secret

Three switches over Enums::PROCESS answered two questions about the same
values: ProcessOutputPanel::processName() gave the command label,
isSensitiveProcess() said whether the output may carry a decrypted entry,
and Pass::emitProcessFinishedSignal() repeated the second one as its own
allow list before dispatching the kind-specific signal. A process kind could
be labelled in the panel and classified as harmless somewhere else.

processinfo.cpp has the table, indexed by PROCESS, and a static_assert that
it covers the enum: a new kind cannot be added without deciding what it is
called and whether its output is a secret. The panel's two functions are
lookups, the dispatch is a hash of signal pointers, and the enum's own
documentation points at the table rather than at three places to remember.

No behaviour change; the labels, the allow list and the signals are what the
existing tests already pinned.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01JuQsrHonihp1nARE7bzstc

* Split the five methods CodeFactor calls complex

Each one did several things in one body, so the branch counts added up:

- reencryptSingleFile() ran decrypt, encrypt, verify, place and commit
  inline; the four steps are named functions now and it reads as the
  sequence it is (cx 26 -> 7).
- reencryptFiles() carried the pull, the entry walk and the per-folder
  recipient check; same three, extracted (cx 21 -> 12).
- GpgIdGeneration::accept() built its refusal texts inline, including a
  lambda for the "how to get through" sentence; the folder binding, that
  sentence and the stale-list refusal are functions of their own
  (cx 27 -> 11). parse() gets its digit grammar out of the loop (cx 22 -> 14).
- Base32::decode() had the padding shapes in a switch and the alphabet in
  the inner loop; both are sm... (continued)

Coverage Stats

214 of 216 new or added lines in 7 files covered. (99.07%)

35 existing lines in 4 files now uncovered.

8788 of 9454 relevant lines covered (92.96%)

150.46 hits per line