Add code duplication analysis to the dev workflow

[yarikoptic]

Add code duplication analysis (jscpd) to the development workflow. It should help keeping my alergy under control! ;)

Changes

• Added .jscpd.json (threshold 0%), ignoring caches/build artifacts and
  CHANGELOG.md (its "duplication" is auto-generated release-note
  boilerplate from intuit/auto, not code to refactor)
• Wired a tox -e duplication check backed by tools/check-duplication.sh
• Added the duplication toxenv to CI (.github/workflows/test.yaml)
• Refactored 4 test files to remove all pre-existing duplication (all of it
  was in test/, none in src/):
  • test_prepare_outputs.py: 16 near-identical tests → 1 parametrized test
  • test_aggregation.py: extracted shared Report-construction/assertion helpers
  • test_report.py: parametrized/extracted shared setup
  • test_ls.py, test_plot.py: extracted small shared helpers
• Threshold tightened to 0% after cleanup (as project grows, accepting some small % is acceptable)

Testing

✅ tox -e duplication passes at 0% duplication
✅ Full test suite passes (471 tests)
✅ tox -e lint / tox -e typing pass on the touched files

---

🤖 Generated with Claude Code using about to be submitted /introduce-duplicates-analysis CON/skill

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 92.26%. Comparing base (7ad5f79) to head (edba47e).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #450      +/-   ##
==========================================
- Coverage   92.41%   92.26%   -0.15%     
==========================================
  Files          15       15              
  Lines        1344     1344              
  Branches      182      182              
==========================================
- Hits         1242     1240       -2     
- Misses         68       69       +1     
- Partials       34       35       +1     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Copilot]

Copilot review overview

🟡 Changes recommended

A critical test assertion is stale, and CI/tooling depend on unpinned or implicit environments.

Review effort: Lite
Findings: 1 · 1

Open (2)

• Refresh current_sample after report updates · New
• Pin the jscpd version for reproducible CI · New

What changed in this PR

Adds zero-threshold jscpd duplication checks to tox and GitHub Actions, with refactoring of duplicated tests.

Changes:

• Adds jscpd configuration and checking tooling.
• Integrates duplication analysis into tox and CI.
• Refactors repetitive tests into parametrized tests and shared helpers.

File	Summary	Review notes
tox.ini	Adds the duplication tox environment.	—
tools/​check-duplication.sh	Runs jscpd analysis.	Pin @latest to a tested version (moderate, 3 votes).
test/​test_plot.py	Extracts shared plotting helpers.	—
test/​test_ls.py	Extracts shared prefix parsing helpers.	—
test/​duct_main/​test_report.py	Consolidates report test setup.	—
test/​duct_main/​test_prepare_outputs.py	Parametrizes repetitive output tests.	—
test/​duct_main/​test_aggregation.py	Extracts aggregation test helpers.	Refresh current_sample after updates (critical, 1 vote).
.jscpd.json	Configures duplication detection and exclusions.	—
.gitignore	Ignores temporary analysis files.	—
.github/​workflows/​test.yaml	Runs duplication checks in CI.	Provision Node.js explicitly (moderate, 1 vote).

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[yarikoptic]

@asmacdo please review and potentially accept this one since it is likely to breed conflicts very easily due to massive pass through the tests

[asmacdo]

Ill review and ill probably cause conflicts with my other tests, which ill fixup here

[yarikoptic-gitmate]

test (macos-15-intel, 3.14, py) failed on edba47e, but not because of this PR's changes:

• The failures are test_e2e.py::test_session_modes[*-new-session] ("No samples collected") and test_execution.py::test_signal_kill[10|-1]. This PR changes neither file nor any src/ code.
• The log shows 28 reruns and a KeyboardInterrupt while a spawned subprocess was still importing stdlib modules. That looks like an overloaded runner rather than wrong behaviour.
• The same tests passed in every other job on this commit, and in this same job on the previous commit (5e520a7), whose source tree is identical.

I'll re-run the failed job once the workflow run finishes; GitHub refuses a re-run while the run is still going. The sampling and signal-handling e2e tests depend on timing, and making them robust on slow macOS runners would be a separate change.

---

Generated by Claude Code

[yarikoptic-gitmate]

Update on CI for edba47e: the macOS 3.14 job passed on re-run, so every test job is now green.

codecov/project still fails at 92.26% (-0.15%). The only lines that differ are the race-only branches in monitor_process (src/con_duct/_tracker.py:258,265). They run only when collect_sample() returns None while the process is still alive, which happens on some runs and not others, and no test targets them on main or here. This PR changes no src/ code: the base and head both have 1344 lines in the same files.

A while ago the same kind of drift hit ProcessStats.aggregate (the [cmd] replacement), and edba47e added a deterministic test for it. The monitor_process race would need a mocked poll()/collect_sample() test, which is outside this PR. Codecov also has no tolerance configured (no codecov.yml), so any drift at all turns the check red.

---

Generated by Claude Code