feat: Phase 2 — leadership 大屏 (/overview) + district normalization + drawer a11y

Phase 2 of the UX modernization. Three conflict-free workstreams.

Leadership 驾驶舱 (/overview):
- Wuhan 13-district Leaflet choropleth (public/wuhan_districts.geojson, keyed
  on name, darker=higher per 高风险高亮), legend, hover/click-zoom
- 全部/门诊/住院 Segmented toggle drives choropleth + Top-5 district bar
- literal "数据截至2023-12" as-of badge (D3 honesty); raw spinner → LoadingState
- decompose OverviewDashboard 501→273; 6 components + 2 helpers under components/overview/

District normalization (backend data boundary):
- case_loader.normalize_district + load_cases_by_district_daily collapse the
  26 dirty labels (武昌/武昌区…) → 13 canonical; analysis/grid/insights repointed
  (fixes a grid-merge row-drop bug as a bonus); in-memory, schema unchanged

Shell a11y (code-review carryover):
- drawer is now a proper modal: ESC, body scroll-lock, focus-in + focus-trap
  cycle + focus-restore, role=dialog/aria-modal/aria-label, hamburger aria-expanded
- SideNav expanded state lifted to AppShell so rail+drawer stay in sync
- RouteErrorBoundary around <Outlet/> keeps shell chrome on page/chunk failure

Gates: tsc 0 · vitest 64 · e2e 19/19 (17 user-flows + 2 overview) · build ok
· backend pytest 6 new + 48 regression green

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

backend/tests/test_district_normalization.py

@@ -0,0 +1,82 @@
+"""Tests for district label normalization at the case-loader boundary.
+
+The processed/cases_by_district_daily.parquet carries both bare ("武昌") and
+区-suffixed ("武昌区") spellings of each district (26 labels = 13 districts × 2
+spellings), which double-counts in any roll-up. data.case_loader normalizes
+these to the canonical 13 区-suffixed names and re-aggregates. These tests pin
+that behavior.
+"""
+import sys
+from pathlib import Path
+
+import pandas as pd
+import pytest
+
+# Ensure the backend package root is importable at collection time (mirrors the
+# sys.path handling other modules rely on once the app is imported).
+BACKEND_ROOT = Path(__file__).parent.parent
+if str(BACKEND_ROOT) not in sys.path:
+    sys.path.insert(0, str(BACKEND_ROOT))
+
+from data.case_loader import (  # noqa: E402
+    CANONICAL_DISTRICTS,
+    normalize_district,
+    load_cases_by_district_daily,
+)
+
+PROJECT_ROOT = Path(__file__).parent.parent.parent
+RAW_PARQUET = PROJECT_ROOT / "processed" / "cases_by_district_daily.parquet"
+
+
+def test_normalize_district_known_bare_forms():
+    """Every known bare form maps to its canonical 区-suffixed name."""
+    cases = {
+        "武昌": "武昌区", "汉阳": "汉阳区", "江岸": "江岸区", "硚口": "硚口区",
+        "青山": "青山区", "洪山": "洪山区", "东西湖": "东西湖区", "汉南": "汉南区",
+        "蔡甸": "蔡甸区", "江夏": "江夏区", "黄陂": "黄陂区", "新洲": "新洲区",
+        "江汉": "江汉区",
+    }
+    for bare, canonical in cases.items():
+        assert normalize_district(bare) == canonical
+
+
+def test_normalize_district_already_suffixed_passes_through():
+    for d in CANONICAL_DISTRICTS:
+        assert normalize_district(d) == d
+
+
+def test_canonical_set_is_exactly_thirteen():
+    assert len(CANONICAL_DISTRICTS) == 13
+    assert len(set(CANONICAL_DISTRICTS)) == 13
+
+
+@pytest.mark.skipif(not RAW_PARQUET.exists(), reason="case parquet not present")
+def test_loader_collapses_to_thirteen_canonical_districts():
+    df = load_cases_by_district_daily()
+    districts = set(df["district"].unique())
+
+    # (a) exactly 13 unique districts, all canonical
+    assert len(districts) == 13, f"expected 13 districts, got {len(districts)}: {sorted(districts)}"
+    assert districts == set(CANONICAL_DISTRICTS)
+
+    # (b) no bare / unsuffixed duplicates remain
+    for name in districts:
+        assert name.endswith(("区", "县", "市")), f"unsuffixed district leaked: {name}"
+
+
+@pytest.mark.skipif(not RAW_PARQUET.exists(), reason="case parquet not present")
+def test_loader_preserves_totals_no_rows_dropped_or_double_counted():
+    """Sum integrity: normalized total == raw parquet total."""
+    raw = pd.read_parquet(RAW_PARQUET)
+    normalized = load_cases_by_district_daily()
+
+    assert int(normalized["total_cases"].sum()) == int(raw["total_cases"].sum())
+    assert int(normalized["outpatient_count"].sum()) == int(raw["outpatient_count"].sum())
+    assert int(normalized["inpatient_count"].sum()) == int(raw["inpatient_count"].sum())
+
+
+@pytest.mark.skipif(not RAW_PARQUET.exists(), reason="case parquet not present")
+def test_raw_parquet_actually_has_dirty_labels():
+    """Sanity: the raw file really has the 26-label problem we are fixing."""
+    raw = pd.read_parquet(RAW_PARQUET)
+    assert raw["district"].nunique() > 13