lekss361
|
fdb54834e5
|
fix(nspd): rename migration 89→98, fix red lines ST_Area→ST_Length (#220)
- data/sql/98_*: rename from 89 to avoid collision with existing 89_drop_dead_brin
- _get_red_lines: ST_Length(ST_Intersection(planar)::geography) instead of
ST_Area(ST_Intersection(::geography, ::geography)) — fixes PostGIS 3.4
tolerance error and returns correct non-zero length for LINESTRING intersections
- Rename intersection_area_sqm → intersection_length_m across schema, TS types,
frontend component, and tests; add test_get_red_lines_db_exception_returns_empty
Addresses review-bot feedback on PR #220.
|
2026-05-16 19:33:14 +03:00 |
|
lekss361
|
ed3c128528
|
feat(nspd): TIER 4 opportunity layers + red lines (#94 PR2 of 4)
- NSPDClient: QUARTER_OPPORTUNITY_LAYERS (auction/scheme/free/future/oopt)
+ QuarterDump.opportunity field
- nspd_sync: harvest_quarter accepts include_opportunity, denorm cols
has_auction_parcels + opportunity_count in UPSERT
- quarter_dump_lookup:
- _get_opportunity_parcels (sort by distance, early-exit on count=0)
- _get_red_lines (query existing dump.red_lines core path, layer='red_lines')
- SQL 89_*: has_auction_parcels + opportunity_count + partial index
- Pydantic: OpportunityParcel + RedLine schemas
- /analyze: nspd_opportunity_parcels + nspd_red_lines fields
- Frontend: NspdOpportunityBlock + NspdRedLinesBlock + LandTab integration
- Tests: 28 total (11 PR1 + 17 PR2), all pass
Red lines uses existing core harvest path (layer 879243 already in dump.red_lines
from PR1+core) — no duplicate harvest, no red_lines_count_v2 column needed.
Part of #94
|
2026-05-16 19:06:22 +03:00 |
|