From 07ec07b7b3900db277157498ee33d294714910cc Mon Sep 17 00:00:00 2001
From: grigo <blacksmoke576@gmail.com>
Date: Tue, 23 Jun 2026 10:39:05 +0300
Subject: [PATCH] added bootstrap buildings

---
 README.md                      |  4 ++-
 db/init/01_postgis.sql         |  1 +
 db/sql/normalize_buildings.sql | 56 ++++++++++++++++++++++++++++++++++
 scripts/load_buildings.sh      | 23 +++++++-------
 4 files changed, 72 insertions(+), 12 deletions(-)
 create mode 100644 db/sql/normalize_buildings.sql

diff --git a/README.md b/README.md
index d6934c1..6becba8 100644
--- a/README.md
+++ b/README.md
@@ -51,7 +51,9 @@ docker compose restart api worker
 ```
 
 The script builds a local `radio-osm2pgsql:latest` image from
-`scripts/Dockerfile.osm2pgsql` on first run.
+`scripts/Dockerfile.osm2pgsql` on first run. It imports the PBF through standard
+`osm2pgsql` tables, then normalizes `planet_osm_polygon` into the API `buildings`
+table with `db/sql/normalize_buildings.sql`.
 
 The import writes a `buildings` table with `geom`, `height_m`, `levels`,
 `building_type`, and `source`. Heights come from `height`, then
diff --git a/db/init/01_postgis.sql b/db/init/01_postgis.sql
index 576e542..67a0700 100644
--- a/db/init/01_postgis.sql
+++ b/db/init/01_postgis.sql
@@ -1 +1,2 @@
 CREATE EXTENSION IF NOT EXISTS postgis;
+CREATE EXTENSION IF NOT EXISTS hstore;
diff --git a/db/sql/normalize_buildings.sql b/db/sql/normalize_buildings.sql
new file mode 100644
index 0000000..84e4919
--- /dev/null
+++ b/db/sql/normalize_buildings.sql
@@ -0,0 +1,56 @@
+DROP TABLE IF EXISTS buildings CASCADE;
+
+CREATE TABLE buildings AS
+WITH raw AS (
+  SELECT
+    osm_id,
+    way,
+    NULLIF(building, '') AS building_type,
+    substring(replace(COALESCE(tags -> 'height', ''), ',', '.') FROM '[-+]?[0-9]+(\.[0-9]+)?')::real
+      AS height_value,
+    substring(replace(COALESCE(tags -> 'building:levels', ''), ',', '.') FROM '[-+]?[0-9]+(\.[0-9]+)?')::real
+      AS levels_value
+  FROM planet_osm_polygon
+  WHERE building IS NOT NULL
+    AND building <> ''
+    AND building <> 'no'
+),
+normalized AS (
+  SELECT
+    osm_id,
+    ST_Multi(ST_CollectionExtract(ST_MakeValid(way), 3)) AS geom,
+    building_type,
+    height_value,
+    levels_value
+  FROM raw
+)
+SELECT
+  osm_id,
+  geom,
+  CASE
+    WHEN height_value IS NOT NULL AND height_value > 0 THEN height_value
+    WHEN levels_value IS NOT NULL AND levels_value > 0 THEN levels_value * 3.0
+    WHEN building_type IN ('garage', 'garages', 'shed') THEN 3.0
+    WHEN building_type IN ('industrial', 'warehouse') THEN 8.0
+    WHEN building_type IN ('church', 'cathedral') THEN 12.0
+    ELSE 9.0
+  END::real AS height_m,
+  CASE
+    WHEN levels_value IS NOT NULL AND levels_value > 0 THEN levels_value::integer
+    ELSE NULL
+  END AS levels,
+  building_type,
+  CASE
+    WHEN height_value IS NOT NULL AND height_value > 0 THEN 'measured'
+    ELSE 'estimated'
+  END AS source
+FROM normalized
+WHERE NOT ST_IsEmpty(geom);
+
+ALTER TABLE buildings ALTER COLUMN geom TYPE geometry(MultiPolygon, 4326)
+  USING ST_SetSRID(geom, 4326);
+
+CREATE INDEX IF NOT EXISTS buildings_geom_gix ON buildings USING GIST (geom);
+CREATE INDEX IF NOT EXISTS buildings_height_idx ON buildings (height_m);
+CREATE INDEX IF NOT EXISTS buildings_type_idx ON buildings (building_type);
+ANALYZE buildings;
diff --git a/scripts/load_buildings.sh b/scripts/load_buildings.sh
index df538d4..263cda9 100644
--- a/scripts/load_buildings.sh
+++ b/scripts/load_buildings.sh
@@ -54,14 +54,20 @@ docker compose exec -T "$POSTGIS_SERVICE" psql -U "$DB_USER" -d "$DB_NAME" \
 
 echo "Drop previous buildings table"
 docker compose exec -T "$POSTGIS_SERVICE" psql -U "$DB_USER" -d "$DB_NAME" \
-  -c "DROP TABLE IF EXISTS buildings CASCADE;"
+  -c "DROP TABLE IF EXISTS buildings CASCADE;
+      DROP TABLE IF EXISTS planet_osm_point CASCADE;
+      DROP TABLE IF EXISTS planet_osm_line CASCADE;
+      DROP TABLE IF EXISTS planet_osm_polygon CASCADE;
+      DROP TABLE IF EXISTS planet_osm_roads CASCADE;
+      DROP TABLE IF EXISTS planet_osm_nodes CASCADE;
+      DROP TABLE IF EXISTS planet_osm_ways CASCADE;
+      DROP TABLE IF EXISTS planet_osm_rels CASCADE;"
 
 echo "Import buildings from $PBF_PATH"
 docker run --rm \
   --network "$NETWORK" \
   -e PGPASSWORD="$DB_PASSWORD" \
   -v "$PBF_DIR:/osm:ro" \
-  -v "$SCRIPT_DIR:/scripts:ro" \
   "$OSM2PGSQL_IMAGE" \
   -H "$POSTGIS_SERVICE" \
   -P 5432 \
@@ -71,16 +77,11 @@ docker run --rm \
   --slim \
   --drop \
   --latlong \
-  --output flex \
-  --style /scripts/osm2pgsql_buildings.lua \
+  --hstore \
   "/osm/$PBF_FILE"
 
-echo "Create buildings indexes"
-docker compose exec -T "$POSTGIS_SERVICE" psql -U "$DB_USER" -d "$DB_NAME" <<'SQL'
-CREATE INDEX IF NOT EXISTS buildings_geom_gix ON buildings USING GIST (geom);
-CREATE INDEX IF NOT EXISTS buildings_height_idx ON buildings (height_m);
-CREATE INDEX IF NOT EXISTS buildings_type_idx ON buildings (building_type);
-ANALYZE buildings;
-SQL
+echo "Normalize buildings table"
+docker compose exec -T "$POSTGIS_SERVICE" psql -U "$DB_USER" -d "$DB_NAME" \
+  < "$REPO_DIR/db/sql/normalize_buildings.sql"
 
 echo "Buildings import complete"