Pure Python WKB (Well Known Binary) Converter

• wkb_to_geojson() converts WKB geometry into GeoJSON
• wkb_to_wkt() converts WKB geometry into WKT
• geojson_to_wkb() converts GeoJSON into WKB
• wkb_to_abstract() converts WKB into an abstract representation which closely resembles the binary format (for debugging purposes)

from parse_wkb import (
    wkb_to_abstract,
    wkb_to_geojson,
    wkb_to_wkt,
    geojson_to_wkb
)

import json

WKB = b"\x01\x04\x00\x00\x00\x04\x00\x00\x00\x01\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00$@\x00\x00\x00\x00\x00\x00D@\x01\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00D@\x00\x00\x00\x00\x00\x00>@\x01\x01\x00\x00\x00\x00\x00\x00\x00\x00\x004@\x00\x00\x00\x00\x00\x004@\x01\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00>@\x00\x00\x00\x00\x00\x00$@"

parsed_to_geojson = wkb_to_geojson(WKB)
assert json.dumps(parsed_to_geojson) == '{"type": "MultiPoint", "coordinates": [[10.0, 40.0], [40.0, 30.0], [20.0, 20.0], [30.0, 10.0]]}'

parsed_to_wkt = wkb_to_wkt(WKB)
assert parsed_to_wkt == 'MULTIPOINT (10.0 40.0, 40.0 30.0, 20.0 20.0, 30.0 10.0)'

parsed_to_abstract = wkb_to_abstract(WKB)
assert parsed_to_abstract == (
	'LEnd',
	('MultiPoint', 'XY'),
	4,
	('LEnd', ('Point', 'XY'), 10.0, 40.0),
	('LEnd', ('Point', 'XY'), 40.0, 30.0),
	('LEnd', ('Point', 'XY'), 20.0, 20.0),
	('LEnd', ('Point', 'XY'), 30.0, 10.0)
)

# the reverse operation is only implemented from GeoJSON at the moment:
encoded_to_wkb_from_geojson = geojson_to_wkb(parsed_to_geojson)
assert encoded_to_wkb_from_geojson == WKB

Supported Geometry Types

Supports POINT, LINESTRING, POLYGON, MULTIPOINT, MULTILINESTRING, MULTIPOLYGON, and GEOMETRYCOLLECTION

Supports geometry with XY, XYZ, XYM, and XYZM dimensions

Specification

This module is based on v1.2.1 of the OpenGIS Implementation Standard for Geographic information - Simple feature access - Part 1: Common architecture (which can be found here https://www.ogc.org/standards/sfa).

Alternate Approaches

This library is somewhat redundant because:

• MySQL, POSTGIS and SQLite (SpatiaLite extension) provide the following conversion functions:

  • ST_AsGeoJSON()
  • ST_GeometryFromGeoJSON()
  • ST_AsWKT()
  • ST_GeometryFromText()

• if you are using geopandas+shapely these libraries provide conversion to and from WKB (implemented with fast C/C++ libraries)

However this library has a limited use-case; when python is working on your system, but you can't get the confounded binaries for geopandas or shapely to compile, or if you don't care for the size of numpy and everything else that comes with a few innocent pip/conda install commands.

From making this script I learned that WKB is actually not a great spec...

• WKB stores a lot of redundant information by repeating the byte order and geometry type for every point.
• Only 8 byte double precision floats are permitted... seems overkill for some applications. Would be better it we could specify precision, or even use integer types.

Planned Features?

• WKT => WKB