OSMRead

Signatures

OSMRead(VARCHAR path);

OSMRead(VARCHAR path, VARCHAR tableName);
OSMRead(VARCHAR path, VARCHAR tableName, BOOLEAN deleteTables);

OSMRead(VARCHAR path, VARCHAR tableName, VARCHAR fileEncoding);
OSMRead(VARCHAR path, VARCHAR tableName, VARCHAR fileEncoding, BOOLEAN deleteTables);

Description

Reads a OSM file from path and creates several tables prefixed by tableName representing the file’s contents. If deleteTables is equal to true, existing tables (with the same prefix) are removed.

11 tables are produced.

  • (1) table_prefix + _tag : table that contains all tag keys listed in the OSM file -> ID_TAG SERIAL PRIMARY KEY, TAG_KEY VARCHAR UNIQUE.

  • (2) table_prefix + _node : table that contains all nodes -> ID_NODE BIGINT PRIMARY KEY, THE_GEOM POINT,ELE DOUBLE PRECISION,USER_NAME VARCHAR,UID BIGINT,VISIBLE BOOLEAN,VERSION INTEGER,CHANGESET INTEGER,LAST_UPDATE TIMESTAMP, NAME VARCHAR.

  • (3) table_prefix + _node_tag : table that contains a list of tag keys (TAG_VALUE) for each node -> ID_NODE BIGINT, ID_TAG BIGINT,TAG_VALUE VARCHAR.

  • (4) table_prefix + _way : table that contains all ways -> ID_WAY BIGINT PRIMARY KEY, USER_NAME VARCHAR, UID BIGINT, VISIBLE BOOLEAN, VERSION INTEGER, CHANGESET INTEGER, LAST_UPDATE TIMESTAMP, NAME VARCHAR.

  • (5) table_prefix + _way_tag : table that contains a list of tag keys (TAG_VALUE) for each way -> ID_WAY BIGINT, ID_TAG BIGINT, VALUE VARCHAR.

  • (6) table_prefix + _way_node : table that contains the list of nodes used to represent a way -> ID_WAY BIGINT, ID_NODE BIGINT, NODE_ORDER INT.

  • (7) table_prefix + _relation: table that contains all relations -> ID_RELATION BIGINT PRIMARY KEY,USER_NAME VARCHAR, UID BIGINT,VISIBLE BOOLEAN,VERSION INTEGER,CHANGESET INTEGER,LAST_UPDATE TIMESTAMP.

  • (8) table_prefix + _relation_tag : table that contains a list of tag keys (TAG_VALUE) for each relation -> ID_RELATION BIGINT, ID_TAG BIGINT, VALUE VARCHAR.

  • (9) table_prefix + _node_member : table that stores all nodes that are referenced into a relation -> ID_RELATION BIGINT,ID_NODE BIGINT, ROLE VARCHAR, NODE_ORDER INT.

  • (10) table_prefix + _way_member : table that stores all ways that are referenced into a relation -> ID_RELATION BIGINT, ID_WAY BIGINT, ROLE VARCHAR, WAY_ORDER INT.

  • (11) table_prefix + _relation_member : table that stores all relations that are referenced into a relation -> ID_RELATION BIGINT, ID_SUB_RELATION BIGINT, ROLE VARCHAR, RELATION_ORDER INT.

By default, the tableName is the filename given in path without the extension.

Warning on the input file name

When a tablename is not specified, special caracters in the input file name are not allowed. The possible caracters are as follow: A to Z, _ and 0 to 9.

The OSM driver supports the following zipped extension : osm.gz and osm.bz2.

Define fileEncoding to force encoding (useful when the header is missing encoding information) (default value is ISO-8859-1).

Examples

1. Using path

CALL OSMRead('/home/user/bretagne.osm');

2. Using path and tableName

CALL OSMRead('/home/user/bretagne.osm', 'OSM_BRETAGNE');

3. Using a zipped file

CALL OSMRead('/home/user/bretagne.osm.gz', 'OSM_BRETAGNE');

4. Using fileEncoding

CALL OSMRead('/home/user/bretagne.osm', 'OSM_BRETAGNE', 'utf-8');

5. Using deleteTables

CALL OSMRead('/home/user/bretagne.osm', 'OSM_BRETAGNE', true);

6. Build OSM data

Based on the created tables, the user can build other OSM data.

H2GIS extract buildings:

DROP TABLE IF EXISTS OSM_BRETAGNE_BUILDINGS;

CREATE TABLE OSM_BRETAGNE_BUILDINGS(ID_WAY BIGINT PRIMARY KEY) AS SELECT DISTINCT ID_WAY FROM OSM_BRETAGNE_WAY_TAG WT, OSM_BRETAGNE_TAG T WHERE WT.ID_TAG = T.ID_TAG AND T.TAG_KEY IN ('building');

DROP TABLE IF EXISTS OSM_BRETAGNE_BUILDINGS_GEOM;

CREATE TABLE OSM_BRETAGNE_BUILDINGS_GEOM AS SELECT ID_WAY, ST_MAKEPOLYGON(ST_MAKELINE(THE_GEOM)) THE_GEOM FROM (SELECT (SELECT ST_ACCUM(THE_GEOM) THE_GEOM FROM (SELECT N.ID_NODE, N.THE_GEOM,WN.ID_WAY IDWAY FROM OSM_BRETAGNE_NODE N,OSM_BRETAGNE_WAY_NODE WN WHERE N.ID_NODE = WN.ID_NODE ORDER BY WN.NODE_ORDER) WHERE  IDWAY = W.ID_WAY) THE_GEOM ,W.ID_WAY FROM OSM_BRETAGNE_WAY W,OSM_BRETAGNE_BUILDINGS B WHERE W.ID_WAY = B.ID_WAY) GEOM_TABLE WHERE ST_GEOMETRYN(THE_GEOM,1) = ST_GEOMETRYN(THE_GEOM, ST_NUMGEOMETRIES(THE_GEOM)) AND ST_NUMGEOMETRIES(THE_GEOM) > 2;

PostGIS extract buildings:

DROP TABLE IF EXISTS OSM_BRETAGNE_BUILDINGS;

CREATE TABLE OSM_BRETAGNE_BUILDINGS AS SELECT DISTINCT ID_WAY FROM OSM_BRETAGNE_WAY_TAG WT, OSM_BRETAGNE_TAG T WHERE WT.ID_TAG = T.ID_TAG AND T.TAG_KEY IN ('building');

ALTER TABLE OSM_BRETAGNE_BUILDINGS ADD PRIMARY KEY(ID_WAY);

DROP TABLE IF EXISTS OSM_BRETAGNE_BUILDINGS_GEOM;

CREATE TABLE OSM_BRETAGNE_BUILDINGS_GEOM AS SELECT ID_WAY,
ST_MAKEPOLYGON(ST_MAKELINE(THE_GEOM)) THE_GEOM FROM (SELECT (SELECT ST_ACCUM(THE_GEOM) THE_GEOM FROM
 (SELECT N.ID_NODE, N.THE_GEOM,WN.ID_WAY IDWAY FROM OSM_BRETAGNE_NODE N,OSM_BRETAGNE_WAY_NODE WN WHERE N.ID_NODE = WN.ID_NODE ORDER BY WN.NODE_ORDER)  ORDEREDNODES WHERE  ORDEREDNODES.IDWAY = W.ID_WAY) THE_GEOM ,W.ID_WAY FROM OSM_BRETAGNE_WAY W,OSM_BRETAGNE_BUILDINGS B WHERE W.ID_WAY = B.ID_WAY) GEOM_TABLE WHERE ARRAY_LENGTH(THE_GEOM, 1) > 2 AND ST_EQUALS(THE_GEOM[1],THE_GEOM[ARRAY_LENGTH(THE_GEOM, 1)]);

See also