improve performance, remove retired addresses and add center on postal

codes

.drone.yml

@@ -11,24 +11,25 @@ steps:
         destination: bestaddresses.zip
     
     - name: convert
-      image: osgeo/gdal:alpine-small-latest
+      image: osgeo/gdal:alpine-normal-latest
+      environment:
+          SELECT: SELECT shape AS geom, id, best_id, coalesce(streetname_fr, streetname_nl, streetname_de) AS streetname, postal_info_objectid, municipality_objectid, COALESCE(municipality_fr, municipality_nl, municipality_de) AS municipality
       commands:
         - unzip -o bestaddresses.zip
-        - ogr2ogr -sql "SELECT DISTINCT COALESCE(municipality_fr, municipality_nl, municipality_de) AS municipality_name, postal_info_objectid, municipality_objectid FROM addresses_complete" postals.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
+        - ogrinfo ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg -sql "CREATE INDEX idx_postal_code ON addresses_complete (postal_info_objectid) WHERE status NOT LIKE 'retired'"
-        - ogr2ogr -sql "SELECT * FROM addresses_complete" -lco GEOMETRY=AS_XY addresses-full.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
+        - ogr2ogr -sql "select St_Centroid(St_Envelope(St_Collect(shape))) AS center, COALESCE(municipality_fr, municipality_nl, municipality_de) AS municipality_name, postal_info_objectid, municipality_objectid FROM addresses_complete where fid % 10 = 1 AND status NOT LIKE 'retired' GROUP BY COALESCE(municipality_fr, municipality_nl, municipality_de), postal_info_objectid, municipality_objectid" -lco GEOMETRY=AS_XY postals.fr.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
-        - ogr2ogr -sql "SELECT * FROM addresses_complete WHERE postal_info_objectid LIKE '1%'" -lco GEOMETRY=AS_XY addresses-1xxx.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
+        - ogr2ogr -sql "$${SELECT} FROM addresses_complete WHERE status NOT LIKE 'retired' LIMIT 10000" -lco GEOMETRY=AS_XY addresses-extract.fr.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
-        - ogr2ogr -sql "SELECT * FROM addresses_complete WHERE postal_info_objectid LIKE '2%'" -lco GEOMETRY=AS_XY addresses-2xxx.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
+        - ogr2ogr -sql "$${SELECT} FROM addresses_complete WHERE status NOT LIKE 'retired'" -lco GEOMETRY=AS_XY addresses-full.fr.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
-        - ogr2ogr -sql "SELECT * FROM addresses_complete WHERE postal_info_objectid LIKE '3%'" -lco GEOMETRY=AS_XY addresses-3xxx.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
+        - ogr2ogr -sql "$${SELECT} FROM addresses_complete WHERE postal_info_objectid LIKE '1%' AND status NOT LIKE 'retired'" -lco GEOMETRY=AS_XY addresses-1xxx.fr.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
-        - ogr2ogr -sql "SELECT * FROM addresses_complete WHERE postal_info_objectid LIKE '4%'" -lco GEOMETRY=AS_XY addresses-4xxx.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
+        - ogr2ogr -sql "$${SELECT} FROM addresses_complete WHERE postal_info_objectid LIKE '2%' AND status NOT LIKE 'retired'" -lco GEOMETRY=AS_XY addresses-2xxx.fr.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
-        - ogr2ogr -sql "SELECT * FROM addresses_complete WHERE postal_info_objectid LIKE '5%'" -lco GEOMETRY=AS_XY addresses-5xxx.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
+        - ogr2ogr -sql "$${SELECT} FROM addresses_complete WHERE postal_info_objectid LIKE '3%' AND status NOT LIKE 'retired'" -lco GEOMETRY=AS_XY addresses-3xxx.fr.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
-        - ogr2ogr -sql "SELECT * FROM addresses_complete WHERE postal_info_objectid LIKE '6%'" -lco GEOMETRY=AS_XY addresses-6xxx.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
+        - ogr2ogr -sql "$${SELECT} FROM addresses_complete WHERE postal_info_objectid LIKE '4%' AND status NOT LIKE 'retired'" -lco GEOMETRY=AS_XY addresses-4xxx.fr.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
-        - ogr2ogr -sql "SELECT * FROM addresses_complete WHERE postal_info_objectid LIKE '7%'" -lco GEOMETRY=AS_XY addresses-7xxx.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
+        - ogr2ogr -sql "$${SELECT} FROM addresses_complete WHERE postal_info_objectid LIKE '5%' AND status NOT LIKE 'retired'" -lco GEOMETRY=AS_XY addresses-5xxx.fr.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
-        - ogr2ogr -sql "SELECT * FROM addresses_complete WHERE postal_info_objectid LIKE '8%'" -lco GEOMETRY=AS_XY addresses-8xxx.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
+        - ogr2ogr -sql "$${SELECT} FROM addresses_complete WHERE postal_info_objectid LIKE '6%' AND status NOT LIKE 'retired'" -lco GEOMETRY=AS_XY addresses-6xxx.fr.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
-        - ogr2ogr -sql "SELECT * FROM addresses_complete WHERE postal_info_objectid LIKE '9%'" -lco GEOMETRY=AS_XY addresses-9xxx.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
+        - ogr2ogr -sql "$${SELECT} FROM addresses_complete WHERE postal_info_objectid LIKE '7%' AND status NOT LIKE 'retired'" -lco GEOMETRY=AS_XY addresses-7xxx.fr.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
+        - ogr2ogr -sql "$${SELECT} FROM addresses_complete WHERE postal_info_objectid LIKE '8%' AND status NOT LIKE 'retired'" -lco GEOMETRY=AS_XY addresses-8xxx.fr.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
+        - ogr2ogr -sql "$${SELECT} FROM addresses_complete WHERE postal_info_objectid LIKE '9%' AND status NOT LIKE 'retired'" -lco GEOMETRY=AS_XY addresses-9xxx.fr.csv ca0fd5c0-8146-11e9-9012-482ae30f98d9_geopackage+sqlite3_3812.gpkg
         - gzip -9 *.csv
-      volumes:
-        - name: mworkdir
-          path: /mworkdir
 
     - name: gitea_release
       image: plugins/gitea-release
@@ -37,9 +38,10 @@ steps:
             from_secret: gitea_key
           title: "v1.0.0"
           files:
-            - "*.csv.gz"
+            - "*.fr.csv.gz"
           base_url: https://gitea.champs-libres.be
-      volumes:
+      trigger:
-        - name: mworkdir
+        event:
-          path: /mworkdir
+          include:
+            - tag
 

.gitignore

@@ -0,0 +1,4 @@
+*.zip
+*.gpkg
+*.csv
+*.qgz

README.md

@@ -2,3 +2,47 @@
 
 Ce dépôt convertit les données d'adresses mises à disposition dans le cadre de [Best Address](https://www.geo.be/catalog/details/ca0fd5c0-8146-11e9-9012-482ae30f98d9?l=fr) en un format csv, utilisable par le logiciel Chill.
 
+
+## Liste des codes postaux
+
+Une liste des codes postaux est générées à partir de la liste des adresses.
+
+### Point pour chaque code postal
+
+Un centre est calculé sur et fixé au centroide de la liste des adresses. Pour plus de rapidité, seule une adresse sur 10 est prise en compte, de manière pseudo-aléatoire sur base de l'identifiant de la ligne.
+
+Justifications de ce filtre pour un code postal (ici, le 4610):
+
+```
+# Calcul du centre sans filtre sur le fid:
+POINT(741703.97364 647537.227319)	4610
+
+# Avec un filtre fid % 5 = 1
+POINT(741669.696698 647591.04102)	4610
+
+# Avec un filtre fid %10 = 1
+POINT(741682.686228 647591.04102)	4610
+```
+
+## Liste des adresses
+
+Les adresses retenues excluent systématiquement celles marquées comme 'retired'.
+
+Les fichiers sont les suivants:
+
+* `addresses-extract`: contient les 10000 premières adresses. A utiliser comme extrait;
+* `addresses-full`: contient l'entièreté des adresses;
+* `addresses-1xxx`: contient les adresses dont le code postal est compris entre 1000 et 1999;
+* `addresses-2xxx`: contient les adresses dont le code postal est compris entre 2000 et 2999;
+* `addresses-3xxx`: contient les adresses dont le code postal est compris entre 3000 et 3999;
+* `addresses-4xxx`: contient les adresses dont le code postal est compris entre 4000 et 4999;
+* `addresses-5xxx`: contient les adresses dont le code postal est compris entre 5000 et 5999;
+* `addresses-6xxx`: contient les adresses dont le code postal est compris entre 6000 et 6999;
+* `addresses-7xxx`: contient les adresses dont le code postal est compris entre 7000 et 7999;
+* `addresses-8xxx`: contient les adresses dont le code postal est compris entre 8000 et 8999;
+* `addresses-9xxx`: contient les adresses dont le code postal est compris entre 9000 et 9999;
+
+## Langue
+
+* Dans les fichiers qui comprennent `fr.csv`, les noms de rues et de municipalités sont choisies d'abord en français si elles sont présentes, ensuite en néerlandais puis en allemand.
+