mirror of https://github.com/Permissionless-Software-Foundation/psf-memo-client.git synced 2026-09-21 16:52:02 -07:00

Files

T

Chris Troutner aaa366a11f Relocate truncate helpers to shared util

Move general-purpose truncateAddr/truncateTxid out of the feed-specific
post-display module into src/util so the recent-profiles view depends on
the shared utility layer rather than a sibling feature module.

By architect.

2026-08-25 16:08:38 -07:00

3.3 KiB

Raw Blame History

Architectural Review Summary — reduce-dry-duplication

Task and commits reviewed

Task: reduce-dry-duplication
Source commit merged and reviewed: eed9d70f6f (Reduce DRY duplication in services and components, from refactorer)
Reviewed as a batch via ready_for_next.sh; merged into swarmforge-architect as a fast-forward and processed.

Architectural findings and fixes applied

Reviewed the refactorer's DRY reduction for UI/Core separation, dependency rule, information hiding/encapsulation, and local code quality.

Findings (mostly sound, two notes):

MemoDb service consolidation (src/services/memo-db.js) — getRecentProfiles/ getRecentPosts → getRecent, and getProfile/getProfilePic/getName → getLevelResource. Preserves error semantics (recent → always throws; level → null on 404). Good cohesion, HTTP details stay in the service. No change needed.
AppUtil pasteFromClipboard (src/util/index.js) — centralizes clipboard-paste into the shared utility; components no longer duplicate clipboard reading. Good information hiding. No change needed.
PlaceholderView extraction — shared placeholder used by placeholder2/3. Cohesive. No change needed.
wallet-summary blur toggle generalization — toggleBlur(field) via dynamic setter name reduces duplication. Functional and correct; the dynamic set${Capitalized} construction is a mild readability tradeoff but acceptable to preserve the DRY intent. No change.
Cross-feature dependency (fixed) — recent-profiles imported the general-purpose truncateAddr/truncateTxid from the feed-specific post-feed/post-display. For cohesion and dependency direction, these pure string helpers belong in the shared utility layer. Applied fix: moved truncateAddr/truncateTxid to src/util/index.js; post-display now imports from util and re-exports them for its existing consumers; recent-profiles imports directly from src/util. This removes a sibling-feature → sibling-feature module coupling and keeps generic display helpers in the shared utility.

Verification results

Mutation (mutate4javascript --scan): memo-db.js 4 sites, util/index.js 6 sites, post-display.js 8 sites. No JS unit test suite exists in this project (npm test → echo 'no tests'); there is no coverage harness, so all sites are uncovered and no survivors can be killed by tests. JSX component files cannot be parsed by the mutation tool (no jsx Babel plugin); they are UI adapters and are outside the currently testable boundary.
DRY (dry4javascript src): no duplicate candidates found — the reduction is effective and my fix introduced no duplication.
Gherkin acceptance mutation (soft): not runnable — this project has no .feature files, no acceptance pipeline, and no runner adapter.
Property tests: none present in this project.

Suite status

No JavaScript unit/acceptance suite exists (npm test echoes "no tests").
bb.edn test task covers the swarmforge helper scripts only (unrelated to this UI work).

Handoffs sent

git_handoff → coder, refactorer (priority 00, task reduce-dry-duplication), to review the architectural commit.

By architect.

3.3 KiB Raw Blame History