Cover the voyage stats in the Stats persistence tests

[dmccoystephenson]

Summary

Stage B (unit-test expansion) cycle targeting Stats and StatsJsonReaderWriter. Only test files are changed; no production code is touched.

• The nine voyage/adventure lifetime totals (totalHaulingContracts, totalTransportRuns, totalRaids, totalPlunder, totalMoneyFromVoyages, crewLostToPiracy, boatsOwned, totalVoyagesCaptained, totalVoyagesFoundered) were written and read by src/stats/statsJsonReaderWriter.py but were not mentioned by any test. A round-trip test (distinct per-field values, float money values, schema-validated) and a legacy-save test (fields deleted → load as 0, other fields kept) are added.
• Two drift guards are added: createJsonFromStats must write exactly the attributes of Stats() and exactly the properties of schemas/stats.json, and createStatsFromJson({}) must reproduce every Stats() default (the existing missing-fields test only checked the seven core fields).
• tests/stats/test_stats.py previously asserted 10 of the 29 defaults; the remaining 19 are now asserted, plus a check that the earnedMilestones/unlockedFeatures lists are not shared between instances.

No tracking issue — the gap was found during triage (no issues were open at triage time, so no issues were deferred).

Test plan

• python3 -m compileall -q src
• Full suite with dummy SDL drivers (run in-process via pytest.main): 985 passed, 1 failed. The single failure, tests/test_saveFileManager.py::test_migrate_old_save_files_skips_a_slot_holding_any_file, reproduces identically on unchanged main in this environment and is caused by the locally installed editable tak differing from the requirements.txt pin; CI on main passes it.
• CI Run Unit Tests green on the PR head

This PR description was drafted during a Gardener session (https://github.com/Stephenson-Software/gardener).

🤖 Generated with Claude Code

---

drafted by Claude on behalf of Daniel Stephenson

[dmccoystephenson]

Self-review rubric (diff: 2 test files, +127 lines, no production code):

• Scope: PASS — only tests/stats/test_stats.py and tests/stats/test_statsJsonReaderWriter.py are modified; both are the mirrors of the Stats/StatsJsonReaderWriter target named in the PR.
• Tests-new: PASS — no new public functions are introduced; six new tests are added.
• Tests-fix: N/A — this is a characterization-only cycle with no bug fix; every new assertion matches current behavior (green locally and in CI).
• Sibling structure: PASS — new tests follow the # prepare / # call / # check layout and reuse the existing createStats / createStatsJsonReaderWriter / getStatsSchema helpers; the legacy-save test mirrors test_save_without_exportStats_loads_with_zeroes.
• Sibling renames: N/A — nothing renamed.
• Docs: PASS — no behavior changed, so README, schemas/stats.json, and PLANNING.md need no update.
• Issue resolution: N/A — no Closes #N; this is a triage-found gap, as stated in the PR body.
• CI: PASS — Run Unit Tests green on the head, log reports 986 passed (980 on main + 6 new).
• Schema-sync: PASS — no persisted field changed; the new test_createJsonFromStats_writesEveryStatsField now enforces writer == Stats() == schema properties.
• Money-format: N/A — no display code touched.
• Deterministic-tests: PASS — no randomness is involved in any new test.
• Headless-pygame: N/A — no pygame code paths.
• camelCase: PASS — new locals (readerWriter, voyageFields, legacyJson) are camelCase.

Judgment call flagged for a human reader:

• tests/stats/test_statsJsonReaderWriter.py — test_createJsonFromStats_writesEveryStatsField asserts exact set equality with schemas/stats.json properties. That is deliberately strict: a schema property kept for a retired field (with no matching Stats attribute) would fail it. No such property exists today; if one is ever wanted, the assertion can be relaxed to a subset check.

Out-of-diff note: locally (not in CI), tests/test_saveFileManager.py::test_migrate_old_save_files_skips_a_slot_holding_any_file fails identically on unchanged main. The cause is the local editable tak differing from the requirements.txt pin; no signal for this diff.

This review comment was drafted during a Gardener session (https://github.com/Stephenson-Software/gardener).

---

drafted by Claude on behalf of Daniel Stephenson