Skip to main content

GeoPackage support from Python. fudgeo is a lightweight package for creating OGC GeoPackages, Feature Classes, and Tables. Easily read and write geometries and attributes to Feature Classes and Tables using regular Python objects and SQLite syntax.

Project description

fudgeo

fudgeo removes the fear uncertainty doubt from using GeoPackages with Python. fudgeo is a lightweight package for creating OGC GeoPackages, Feature Classes, and Tables. Easily read and write geometries and attributes to Feature Classes and Tables using regular Python objects and SQLite syntax.

For details on OGC GeoPackages, please see the OGC web page.

Installation

fudgeo is available from the Python Package Index.

Python Compatibility

The fudgeo library is compatible with Python 3.7 to 3.12. Developed and tested on macOS and Windows, should be fine on Linux too.

Usage

fudgeo can be used to:

  • Create a new empty GeoPackage or open an existing GeoPackage
  • Create new FeatureClass or Table with optional overwrite
  • Create SpatialReferenceSystem for a FeatureClass
  • Build geometry objects from lists of coordinate values
  • Work with data in Table or FeatureClass in a normal SQLite manner (e.g. SELECT, INSERT, UPDATE, DELETE)
  • Retrieve fields from a FeatureClass or Table
  • Access primary key field of FeatureClass or Table
  • Access geometry column name and geometry type for FeatureClass
  • Add spatial index on FeatureClass
  • Drop FeatureClass or Table
  • Add metadata and schema details

Create an Empty GeoPackage / Open GeoPackage

from fudgeo.geopkg import GeoPackage

# Creates an empty geopackage
gpkg: GeoPackage = GeoPackage.create('../data/example.gpkg')

# Opens an existing Geopackage (no validation)
gpkg: GeoPackage = GeoPackage('../data/example.gpkg')

GeoPackages are created with three default Spatial References defined automatically, a pair of Spatial References to handle undefined cases, and a WGS 84 entry.

The definition of the WGS84 entry is flexible - meaning that the WKT for WGS84 can be setup per the users liking. As an example, use with Esri's ArcGIS means either using the EPSG WKT or the ESRI WKT. By default the ESRI WKT is used - However, if EPSG WKT is desired, you may provide a flavor parameter to the create method specifying EPSG.

from fudgeo.geopkg import GeoPackage

# Creates an empty geopackage using EPSG definitions
gpkg: GeoPackage = GeoPackage.create('../temp/test.gpkg', flavor='EPSG')

Create a Feature Class

Use the create_feature_class method of a GeoPackage to make a new feature class. Feature classes require a name and a Spatial Reference, the name must follow SQLite naming requirements. Each feature class is defined with fid and SHAPE fields, additional fields can be defined during creation.

A Feature Class can be created with Z or M (or both) enabled. If either of these options are enabled, the geometry inserted into the Feature Class must include a value for the option specified.

from typing import Tuple
from fudgeo.enumeration import GeometryType, SQLFieldType
from fudgeo.geopkg import FeatureClass, Field, GeoPackage, SpatialReferenceSystem

SRS_WKT: str = (
    'PROJCS["WGS_1984_UTM_Zone_23N",'
    'GEOGCS["GCS_WGS_1984",'
    'DATUM["D_WGS_1984",'
    'SPHEROID["WGS_1984",6378137.0,298.257223563]],'
    'PRIMEM["Greenwich",0.0],'
    'UNIT["Degree",0.0174532925199433]],'
    'PROJECTION["Transverse_Mercator"],'
    'PARAMETER["False_Easting",500000.0],'
    'PARAMETER["False_Northing",0.0],'
    'PARAMETER["Central_Meridian",-45.0],'
    'PARAMETER["Scale_Factor",0.9996],'
    'PARAMETER["Latitude_Of_Origin",0.0],'
    'UNIT["Meter",1.0]]')

SRS: SpatialReferenceSystem = SpatialReferenceSystem(
    name='WGS_1984_UTM_Zone_23N', organization='EPSG',
    org_coord_sys_id=32623, definition=SRS_WKT)
fields: Tuple[Field, ...] = (
    Field('road_id', SQLFieldType.integer),
    Field('name', SQLFieldType.text, size=100),
    Field('begin_easting', SQLFieldType.double),
    Field('begin_northing', SQLFieldType.double),
    Field('end_easting', SQLFieldType.double),
    Field('end_northing', SQLFieldType.double),
    Field('begin_longitude', SQLFieldType.double),
    Field('begin_latitude', SQLFieldType.double),
    Field('end_longitude', SQLFieldType.double),
    Field('end_latitude', SQLFieldType.double),
    Field('is_one_way', SQLFieldType.boolean))

gpkg: GeoPackage = GeoPackage.create('../temp/test.gpkg')
fc: FeatureClass = gpkg.create_feature_class(
    'road_l', srs=SRS, fields=fields, shape_type=GeometryType.linestring,
    m_enabled=True, overwrite=True, spatial_index=True)

About Spatial References For GeoPackages

Spatial References in GeoPackages can use any definition from any authority - be that EPSG, ESRI, or another authority. fudgeo imposes no restriction and performs no checks on the definitions provided. Take care to ensure that the definitions are compatible with the platform / software you intend to utilize with the GeoPackage.

Insert Features into a Feature Class (SQL)

Features can be inserted into a Feature Class using SQL.

This example shows the creation of a random point Feature Class and builds upon the code from previous examples. Note that the create Feature Class portion of the code is omitted...

from random import choice, randint
from string import ascii_uppercase, digits
from typing import List, Tuple

from fudgeo.geometry import LineStringM
from fudgeo.geopkg import GeoPackage

# Generate some random points and attributes
rows: List[Tuple[LineStringM, int, str, float, float, float, float, bool]] = []
for i in range(10000):
    name = ''.join(choice(ascii_uppercase + digits) for _ in range(10))
    road_id = randint(0, 1000)
    eastings = [randint(300000, 600000) for _ in range(20)]
    northings = [randint(1, 100000) for _ in range(20)]
    coords = [(x, y, m) for m, (x, y) in enumerate(zip(eastings, northings))]
    road = LineStringM(coords, srs_id=32623)
    rows.append((road, road_id, name, eastings[0], northings[0],
                 eastings[-1], northings[-1], False))

# NOTE Builds from previous examples
gpkg: GeoPackage = GeoPackage('../data/example.gpkg')   
with gpkg.connection as conn:
    conn.executemany("""
        INSERT INTO road_l (SHAPE, road_id, name, begin_easting, begin_northing, 
                            end_easting, end_northing, is_one_way) 
           VALUES (?, ?, ?, ?, ?, ?, ?, ?)""", rows)

Geometry Examples

Review the tests for fudgeo for a comprehensive look into creating geometries, below are some examples showing the simplicity of this package.

from typing import List, Tuple
from fudgeo.geometry import LineStringZM, Point, Polygon

# Point in WGS 84
pt: Point = Point(x=-119, y=34)

# Line with ZM Values for use with UTM Zone 23N (WGS 84)
coords: List[Tuple[float, float, float, float]] = [
    (300000, 1, 10, 0), (300000, 4000000, 20, 1000),
    (700000, 4000000, 30, 2000), (700000, 1, 40, 3000)]
line: LineStringZM = LineStringZM(coords, srs_id=32623)

# list of rings where a ring is simply the list of points it contains.
rings: List[List[Tuple[float, float]]] = [
    [(300000, 1), (300000, 4000000), (700000, 4000000), (700000, 1), (300000, 1)]]
poly: Polygon = Polygon(rings, srs_id=32623)

Select Features from GeoPackage (SQL)

When selecting features from a GeoPackage feature class use SQL. For the most part (mainly simple geometries e.g. those without Z or M) this can be done via a basic SELECT statement like:

gpkg = GeoPackage(...)
cursor = gpkg.connection.execute("""SELECT SHAPE, example_id FROM point_fc""")
features = cursor.fetchall()

This will return a list of tuples where each tuple contains a Point object and an integer for example_id field.

When working with extended geometry types (those with Z and/or M) then the approach is to ensure SQLite knows how to convert the geopackage stored geometry to a fudgeo geometry, this is done like so:

from typing import List, Tuple
from fudgeo.geometry import LineStringM
from fudgeo.geopkg import GeoPackage

gpkg: GeoPackage = GeoPackage('../data/example.gpkg')
cursor = gpkg.connection.execute(
    """SELECT SHAPE "[LineStringM]", road_id FROM test""")
features: List[Tuple[LineStringM, int]] = cursor.fetchall()

or a little more general, accounting for extended geometry types and possibility of the geometry column being something other tha SHAPE:

from typing import List, Tuple
from fudgeo.geometry import LineStringM
from fudgeo.geopkg import FeatureClass, GeoPackage

gpkg: GeoPackage = GeoPackage('../data/example.gpkg')
fc: FeatureClass = FeatureClass(geopackage=gpkg, name='road_l')
cursor = gpkg.connection.execute(f"""
    SELECT {fc.geometry_column_name} "[{fc.geometry_type}]", road_id 
    FROM {fc.escaped_name}""")
features: List[Tuple[LineStringM, int]] = cursor.fetchall()

Extensions

Spatial Index Extension

Spatial Index Extension implementation based on section F.3. RTree Spatial Indexes of the GeoPackage Encoding Standard.

Spatial Indexes apply to individual feature classes. A spatial index can be added at create time or added on an existing feature class.

from typing import Tuple
from fudgeo.enumeration import SQLFieldType
from fudgeo.geopkg import FeatureClass, Field, GeoPackage, SpatialReferenceSystem


SRS_WKT: str = (
    'PROJCS["WGS_1984_UTM_Zone_23N",'
    'GEOGCS["GCS_WGS_1984",'
    'DATUM["D_WGS_1984",'
    'SPHEROID["WGS_1984",6378137.0,298.257223563]],'
    'PRIMEM["Greenwich",0.0],'
    'UNIT["Degree",0.0174532925199433]],'
    'PROJECTION["Transverse_Mercator"],'
    'PARAMETER["False_Easting",500000.0],'
    'PARAMETER["False_Northing",0.0],'
    'PARAMETER["Central_Meridian",-45.0],'
    'PARAMETER["Scale_Factor",0.9996],'
    'PARAMETER["Latitude_Of_Origin",0.0],'
    'UNIT["Meter",1.0]]')
SRS: SpatialReferenceSystem = SpatialReferenceSystem(
    name='WGS_1984_UTM_Zone_23N', organization='EPSG',
    org_coord_sys_id=32623, definition=SRS_WKT)
fields: Tuple[Field, ...] = (
    Field('id', SQLFieldType.integer),
    Field('name', SQLFieldType.text, size=100))

gpkg: GeoPackage = GeoPackage.create('../temp/spatial_index.gpkg')
# add spatial index at create time
event: FeatureClass = gpkg.create_feature_class(
    'event_p', srs=SRS, fields=fields, spatial_index=True)
assert event.has_spatial_index is True

# add spatial index on an existing feature class / post create
signs: FeatureClass = gpkg.create_feature_class(
    'signs_p', srs=SRS, fields=fields)
# no spatial index
assert signs.has_spatial_index is False
signs.add_spatial_index()
# spatial index now present
assert signs.has_spatial_index is True

Refer to SQLite documentation on how to use these indexes for faster filtering / querying. Also note how to handle round off error when querying.

Metadata Extension

Metadata Extension implementation based on F.8. Metadata of the GeoPackage Encoding Standard.

The metadata extension is enabled at the GeoPackage level applying to all tables and feature classes. That said, not every table and feature class is required to have metadata.

Metadata extension can be enabled at create time for a GeoPackage or can be enabled on an existing GeoPackage.

from fudgeo.geopkg import GeoPackage

# enable metadata at create time
gpkg: GeoPackage = GeoPackage.create('../data/metadata.gpkg', enable_metadata=True)
assert gpkg.is_metadata_enabled is True

# enable metadata on an existing GeoPackage
gpkg: GeoPackage = GeoPackage('../data/example.gpkg')
assert gpkg.is_metadata_enabled is False
gpkg.enable_metadata_extension()
assert gpkg.is_metadata_enabled is True
from fudgeo.enumeration import MetadataScope
from fudgeo.extension.metadata import TableReference
from fudgeo.geopkg import GeoPackage

# open GeoPackage with metadata extension enabled
gpkg: GeoPackage = GeoPackage('../data/example.gpkg')

# open a metadata xml file and add it to the GeoPackage
with open(...) as fin:
    id_ = gpkg.metadata.add_metadata(
        uri='https://www.isotc211.org/2005/gmd',
        scope=MetadataScope.dataset, metadata=fin.read()
    )
# apply the metadata to a feature class
reference = TableReference(table_name='road_l', file_id=id_)
gpkg.metadata.add_references(reference)

Support provided for the following reference types:

  • GeoPackageReference -- used for GeoPackage level metadata
  • TableReference -- used for Table and FeatureClass level metadata
  • ColumnReference -- used for a column in a Table or FeatureClass
  • RowReference -- used for a row in a Table or FeatureClass
  • RowColumnReference -- used for row / column combination in a Table or FeatureClass

Schema Extension

Schema Extension implementation based on F.9. Schema of the GeoPackage Encoding Standard.

The schema extension is enabled at the GeoPackage level and allows for extended definitions on column names (e.g. name, title, description) and for constraints to be defined for columns. Constraints definitions are intended for applications usage and, while similar, are not the same as database constraints.

Schema extension can be enabled at create time for a GeoPackage or can be enabled on an existing GeoPackage.

from fudgeo.geopkg import GeoPackage

# enable schema at create time
gpkg: GeoPackage = GeoPackage.create('../data/schema.gpkg', enable_schema=True)
assert gpkg.is_schema_enabled is True

# enable schema on an existing GeoPackage
gpkg: GeoPackage = GeoPackage('../data/example.gpkg')
assert gpkg.is_schema_enabled is False
gpkg.enable_schema_extension()
assert gpkg.is_schema_enabled is True
from fudgeo.extension.schema import (
    EnumerationConstraint, GlobConstraint, RangeConstraint)
from fudgeo.geopkg import GeoPackage

# open GeoPackage with schema extension enabled
gpkg: GeoPackage = GeoPackage('../data/example.gpkg')

# add constraints for use with column definitions
constraints = [
    EnumerationConstraint(name='odds', values=[1, 3, 5, 7, 9]),
    EnumerationConstraint(name='colors', values=['red', 'yellow', 'blue']),
    GlobConstraint(name='pin', pattern='[0-9][0-9][0-9][0-9]'),
    RangeConstraint(name='exertion', min_value=6, max_value=20),
    RangeConstraint(name='longitude', min_value=-180, max_value=180),
    RangeConstraint(name='latitude', min_value=90, max_value=90),
]
gpkg.schema.add_constraints(constraints)

# use constrains and set some additional details for column name
gpkg.schema.add_column_definition(
    table_name='road_l', column_name='begin_longitude', 
    name='Beginning Longitude for Road', title='Begin Longitude', 
    constraint_name='longitude')
gpkg.schema.add_column_definition(
    table_name='road_l', column_name='begin_latitude', 
    name='Beginning Latitude for Road', title='Begin Latitude', 
    constraint_name='latitude')
gpkg.schema.add_column_definition(
    table_name='road_l', column_name='end_longitude', 
    name='Ending Longitude for Road', title='End Longitude', 
    constraint_name='longitude')
gpkg.schema.add_column_definition(
    table_name='road_l', column_name='end_latitude', 
    name='Ending Latitude for Road', title='End Latitude', 
    constraint_name='latitude')

Support provided for the following constraint types:

  • EnumerationConstraint -- restrict to one or more values
  • GlobConstraint -- pattern match based constraint
  • RangeConstraint -- value constrained within a range, optionally including the bounds

License

MIT

Release History

v0.7.1

  • ensure support for Python 3.12 and update documentation / configuration

v0.7.0

  • add support for schema extension
  • add support for metadata extension
  • add __geo_interface__ to geometry classes
  • introduce bounding_box property on Envelope class
  • introduce as_tuple method on Point classes
  • add extension sub-package, move spatial module into extension
  • add spatial_index_name property on FeatureClass, returns the index table name
  • enable enforcement of foreign key constraints
  • reorganize code to handle OGR contents like an extension
  • move protected functions from geopkg module into util module and rename
  • add type hinting to enumerations module
  • move EnvelopeCode into enumerations

v0.6.0

  • change ogr_contents default value to False (breaking change)
  • add spatial_index option to FeatureClass creation, default to False
  • add add_spatial_index method to FeatureClass for adding spatial index post creation
  • add has_spatial_index property to FeatureClass
  • add count property to Table and FeatureClass
  • add primary_key_field property to Table and FeatureClass
  • small speed-up to Point unpacking
  • update is_empty to rely on internal attribute data type
  • improvements to SQL statements to handle names that must be escaped
  • bump user_version to reflect adopted version 1.3.1 of OGC GeoPackage
  • add optional views for geometry columns and spatial references

v0.5.2

  • store empty state on the instance during geometry read
  • introduce base classes for common capability and parametrize via class attributes
  • add stub files to provide type hinting specialization

v0.5.1

  • small performance improvements by reducing bytes concatenation and building up bytearray

v0.5.0

  • performance improvements for geometry reading (especially for geometries with large numbers of points / parts)
  • performance improvements for geometry writing
  • incorporated numpy and bottleneck as dependencies

v0.4.2

  • only unpack header and delay unpacking coordinates until needed
  • write envelope to geometry header

v0.4.1

  • unpack envelope from header (when available)
  • add envelope property to 2D, 3D, and 4D geometry
  • derive envelope from underlying coordinates / geometries if not set from header

v0.4.0

  • add string representations to GeoPackage, Table, and FeatureClass
  • allow optional creation of the gpkg_ogr_contents, defaults to True (create)
  • split geometry module into a sub-package

v0.3.10

  • add escaped_name property to BaseTable, applies to Table and FeatureClass
  • escape the name of input table / feature class during create

v0.3.9

  • quote reversal, doubles inside of singles for escaped_name

v0.3.8

  • add fields property to BaseTable, applies to Table and FeatureClass
  • add field_names property to BaseTable, applies to Table and FeatureClass
  • add escaped_name property to Field to return name valid for use in queries
  • add type hinting to embedded sql statements and supporting values

v0.3.7

  • add is_empty property to geometries, improved handling for empty geometries
  • update user_version to 10300
  • add handling for geometry headers with envelopes (skipping content)
  • add type hinting to constants

v0.3.6

  • add srs_id (optional) to SpatialReferenceSystem instantiation, default to org_coord_sys_id if not specified

v0.3.5

  • store coordinates in attribute on 2D, 3D, and 4D geometry
  • avoid creating points on instantiation of geometry
  • expose points property to return point objects for 2D, 3D, and 4D geometry

v0.3.4

  • add from_tuple class methods to Point, PointZ, PointM, and PointZM

v0.3.3

  • catch possible exception when parsing microseconds from time
  • add converter for timestamp to use same converter as datetime
  • use lower case table names in queries

v0.3.2

  • include PolygonM, PolygonZM, MultiPolygonM, and MultiPolygonZM in geometry registration

v0.3.1

  • delay opening a GeoPackage connection until connection property is accessed

v0.3.0

  • add support for PolygonM, PolygonZM, MultiPolygonM, and MultiPolygonZM
  • add geometry_column_name and geometry_type properties to FeatureClass
  • simplify query used by has_z and has_m

v0.2.1

  • improve _convert_datetime to handle different formats for timestamp (contributed by @alexeygribko)

v0.2.0

  • improve _convert_datetime to handle timezone
  • add DATETIME tp SQLFieldType

v0.1.2

  • add option to overwrite feature classes and tables in create_feature_class and create_table methods
  • add option to overwrite in create method on FeatureClass and Table classes
  • add drop method on FeatureClass and Table classes

v0.1.1

  • make compatible with Python 3.7 and up (update type hints, remove walrus)
  • add support for OGR contents table (gpkg_ogr_contents) and triggers
  • add tables and feature_classes properties to GeoPackage class
  • include application_id and user_version in SQL definition
  • fix timestamp format (was missing seconds)

v0.1.0

  • initial release, basic port of legacy pygeopkg package

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distributions

No source distribution files available for this release.See tutorial on generating distribution archives.

Built Distribution

fudgeo-0.7.1-py3-none-any.whl (45.6 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page