def generate_full_geocode_set(cls, nominatim_host, verbose=False):
from shapely.geometry import shape as Shape, mapping as Mapping
from shapely.ops import cascaded_union
country = cls.regenerate_country_geocode(nominatim_host, verbose)
country = {
code: {
'geojson': Mapping(Shape(geo['geojson']).simplify(0.005)),
'display_name': geo['display_name'],
'namedetails': geo.get('namedetails', {})
}
for code, geo in country.iteritems()
if geo.get('geojson') and geo.get('display_name')
}
with open('cgeo.json', 'wb') as out:
dump(country, out)
if exists('cgeo.json.xz'):
if isfile('cgeo.json.xz'):
remove('cgeo.json.xz')
else:
raise TypeError("cgeo.json.xz exists and is not a file!")
Command('xz')('-z9', 'cgeo.json')
regions = cls.regenerate_region_geocode(nominatim_host, verbose)
regions = {
ccode: {
rcode: {
'geojson': Mapping(Shape(geo['geojson']).simplify(0.01)),
'display_name': geo['display_name'],
'namedetails': geo.get('namedetails', {})
}
for rcode, geo in region.iteritems()
if geo.get('geojson') and geo.get('display_name')
}
for ccode, region in regions.iteritems()
}
with open('rgeo.json', 'wb') as out:
dump(regions, out)
if exists('rgeo.json.xz'):
if isfile('rgeo.json.xz'):
remove('rgeo.json.xz')
else:
raise TypeError("rgeo.json.xz exists and is not a file!")
Command('xz')('-z9', 'rgeo.json')
phones = cls.regenerate_phone_geocode(nominatim_host, verbose)
phones = {
ccode: {
acode: cascaded_union(map(Shape, phones[ccode][acode])
).centroid.coords[0][::-1]
for acode in phones[ccode]
}
for ccode in phones
}
with open('pgeo.json', 'wb') as out:
dump(phones, out)
if exists('pgeo.json.xz'):
if isfile('pgeo.json.xz'):
remove('pgeo.json.xz')
else:
raise TypeError("pgeo.json.xz exists and is not a file!")
Command('xz')('-z9', 'pgeo.json')
# Filter for the selected region.
plzs = {
plz
for plz, bundesland in mapping_plz_bundesland.items()
if bundesland == configuration.RESTRICT_REGION
}
# Only keep the geometries for the selected region.
# Afterwards cache the shape data to avoid repeated conversions.
geometries = shp.shapeRecords()
geometries_filtered = [
geometry for geometry in geometries if geometry.record[0] in plzs
]
for geometry in geometries_filtered:
setattr(geometry, "boundary", Shape(geometry.shape))
def is_in_region(lat, lon):
point = Point(lon, lat)
for geometry in geometries_filtered:
if point.within(geometry.boundary):
return True
return False
else:
# No restrictions requested, so provide a dummy implementation.
def is_in_region(lat, lon):
return True
def init():
global is_done_loading
def _admin_cc(filename):
parts = path.basename(filename).split('.', 1)[0].split('-')
admin, cc = parts if len(parts) == 2 else (parts[0], None)
admin = int(admin[-1])
return admin, cc
nearest_points = {0: [], 1: [], 2: []}
shapes = {0: [], 1: [], 2: []}
cc_shapes = {}
id_regions = {}
us_states = {}
files = glob(path.join(GEOJSON_DIR, 'admin*.json'))
if not files:
is_done_loading = RuntimeError(
'Missing GeoJSON files. '
'In development environments, merge in the "json"'
'branch. See CONTRIBUTING.md')
raise is_done_loading
log.debug('Loading GeoJSON data ...')
for filename in files:
admin, cc = _admin_cc(filename)
with open(filename, encoding='utf-8') as f:
collection = json.load(f, encoding='utf-8')
for feature in collection['features']:
p = feature['properties']
shape = Shape(feature['geometry'])
tup = (shape, p)
# Add representative lat-lon pair if non-existent
if (np.isnan(p.get('latitude', np.nan)) or
np.isnan(p.get('longitude', np.nan))):
try:
r = shape.representative_point()
except ValueError:
# For GBR, representative point above fails with:
# ValueError: No Shapely geometry can be created from null value
r = shape.centroid
if not r.within(shape):
max_poly = max([shape] if isinstance(shape, Polygon) else list(shape),
key=lambda polygon: polygon.area)
# From https://stackoverflow.com/questions/33311616/find-coordinate-of-closest-point-on-polygon-shapely/33324058#33324058
poly_ext = max_poly.exterior
dist = poly_ext.project(r)
pt = poly_ext.interpolate(dist)
r = Point(pt.coords[0])
assert r.within(shape)
p.update(latitude=r.y, longitude=r.x)
# Get representative points for the k-d tree
points = []
polygons = (shape,) if isinstance(shape, Polygon) else shape
for poly in polygons:
points.append([poly.centroid.y, poly.centroid.x])
if poly.area > 10:
points.extend(xy[::-1]
for xy in poly.simplify(1).exterior.coords)
nearest_points[admin].extend(points)
tups = [tup] * len(points)
shapes[admin].extend(tups)
id_regions[p['_id']] = tup
if admin == 0:
cc_shapes[p['adm0_a3']] = tup
# Make Admin1 shapes available in Admin2 too. Except for USA
# which is the country we have explicit Admin2 shapes for
if admin == 1 and cc not in ADMIN2_COUNTRIES:
shapes[2].extend(tups)
nearest_points[2].extend(points)
if admin == 1 and cc == 'USA':
us_states[p['hasc'].split('.')[1]] = tup
kdtree = {admin: KDTree(np.array(centroids))
for admin, centroids in nearest_points.items()}
cc_shapes['NUL'] = (None, NUL) # tuple for Null Island
assert all(len(nearest_points[admin]) == len(shapes[admin])
for admin in shapes)
global SHAPES, CC_SHAPES, KDTREE, ID_REGIONS, US_STATES
SHAPES, CC_SHAPES, KDTREE, ID_REGIONS, US_STATES = shapes, cc_shapes, kdtree, id_regions, us_states
is_done_loading = True
return
def is_point_within_shape(x: float, y: float, geometry):
point = Point(x, y)
shape = Shape(geometry)
return shape.contains(point)