def raster_to_metadata(
raster: Union[List[Union[str, gdal.Dataset]], str, gdal.Dataset],
create_geometry: bool = False,
) -> Union[Metadata_raster, List[Metadata_raster]]:
"""Reads a raster from a list of rasters, string or a dataset and returns metadata.
Args:
raster (list, path | Dataset): The raster to calculate metadata for.
**kwargs:
create_geometry (bool): If False footprints of the raster is calculated, including
in latlng (wgs84). Requires a reprojection check. Do not use if not required
and performance is essential. Produces geojsons as well.
Returns:
A dictionary containing metadata about the raster.
"""
type_check(raster, [str, gdal.Dataset], "raster")
type_check(create_geometry, [bool], "create_geometry")
input_list = get_raster_path(raster, return_list=True)
return_list = []
for readied_raster in input_list:
if is_raster(readied_raster):
return_list.append(
_raster_to_metadata(readied_raster,
create_geometry=create_geometry))
else:
raise TypeError(f"Input: {readied_raster} is not a raster.")
if isinstance(raster, list):
return return_list
return return_list[0]
def get_vector_path(vector: Union[str, ogr.DataSource]) -> str:
"""Takes a string or a ogr.Datasource and returns its path.
Args:
vector (path | Dataset): A path to a vector or an ogr.Datasource.
Returns:
A string representing the path to the vector.
"""
if isinstance(vector, str) and (len(vector) >= 8
and vector[0:8]) == "/vsimem/":
return vector
type_check(vector, [str, ogr.DataSource], "vector")
opened = open_vector(vector, convert_mem_driver=True, writeable=False)
try:
path = str(opened.GetDescription())
if len(path) >= 8 and path[0:8] == "/vsimem/":
return path
elif os.path.exists(path):
return path
else:
raise Exception(f"Error while getting path from raster: {vector}")
except:
raise Exception(f"Error while getting path from raster: {vector}")
def get_raster_path(raster: Union[list[str, gdal.Dataset], str, gdal.Dataset],
return_list=False) -> str:
"""Takes a string or a gdal.Dataset and returns its path.
Args:
raster (list | path | Dataset): A path to a raster or a GDAL dataframe.
return_list (bool): If True, returns a list of paths.
Returns:
A string representing the path to the raster.
"""
type_check(raster, [list, str, gdal.Dataset], "raster")
return_list = []
if isinstance(raster, list):
input_list = raster
else:
input_list = [raster]
for readied_raster in input_list:
return_list.append(_get_raster_path(readied_raster))
if isinstance(raster, list) or return_list:
return return_list
return return_list[0]
def vector_feature_to_layer(
vector: Union[List[Union[str, ogr.DataSource]], str, ogr.DataSource],
fid: int,
layer: int = 1,
) -> List[str]:
"""Adds a spatial index to the vector if it doesn't have one.
Args:
vector (list, path | vector): The vector to add the index to.
Returns:
A path to the original vector.
"""
type_check(vector, [list, str, ogr.DataSource], "vector")
vector_list = to_vector_list(vector)
try:
for in_vector in vector_list:
metadata = internal_vector_to_metadata(in_vector)
ref = open_vector(in_vector)
for layer in metadata["layers"]:
name = layer["layer_name"]
geom = layer["column_geom"]
sql = f"SELECT CreateSpatialIndex('{name}', '{geom}') WHERE NOT EXISTS (SELECT HasSpatialIndex('{name}', '{geom}'));"
ref.ExecuteSQL(sql, dialect="SQLITE")
except:
raise Exception(f"Error while creating indices for {vector}")
return vector_list
def singlepart_to_multipart(
vector: Union[List[Union[str, ogr.DataSource]], Union[str, ogr.DataSource]],
out_path: str = None,
overwrite: bool = True,
add_index: bool = True,
process_layer: int = -1,
) -> Union[List[str], str]:
type_check(vector, [list, str, ogr.DataSource], "vector")
type_check(out_path, [str], "out_path", allow_none=True)
type_check(overwrite, [bool], "overwrite")
type_check(add_index, [bool], "add_index")
type_check(process_layer, [int], "process_layer")
vector_list, path_list = ready_io_vector(vector, out_path, overwrite=overwrite)
output = []
for index, in_vector in enumerate(vector_list):
output.append(
internal_singlepart_to_multipart(
in_vector,
out_path=path_list[index],
overwrite=overwrite,
add_index=add_index,
process_layer=process_layer,
)
)
if isinstance(vector, list):
return output[0]
return output
def raster_get_nodata_value(
raster: Union[List[Union[gdal.Dataset, str]], gdal.Dataset, str],
) -> Union[List[Optional[Number]], Optional[Number]]:
"""Get the nodata value of a raster or a from a list of rasters.
Args:
raster (path | raster | list): The raster(s) to retrieve nodata values from.
Returns:
Returns the nodata value from a raster or a list of rasters
"""
type_check(raster, [list, str, gdal.Dataset], "raster")
rasters = get_raster_path(raster, return_list=True)
nodata_values = []
for internal_raster in rasters:
if not is_raster(internal_raster):
raise ValueError(f"Input raster is invalid: {internal_raster}")
raster_metadata = raster_to_metadata(internal_raster)
if not isinstance(raster_metadata, dict):
raise Exception("Metadata is in the wrong format.")
raster_nodata = raster_metadata["nodata_value"]
nodata_values.append(raster_nodata)
if isinstance(raster, list):
return nodata_values
else:
return nodata_values[0]
def is_raster(raster):
"""Checks if a raster is valid.
Args:
raster (str | gdal.Dataset): A path to a raster or a GDAL dataframe.
Returns:
A boolean.
"""
type_check(raster, [str, gdal.Dataset], "raster")
if isinstance(raster, str):
if not file_exists(raster) and not path_is_in_memory(raster):
return False
try:
opened = gdal.Open(raster, 0)
except:
return False
if opened is None:
return False
return True
if isinstance(raster, gdal.Dataset):
return True
return False
def reproject_vector(
vector: Union[List[Union[str, ogr.DataSource]], Union[str,
ogr.DataSource]],
projection: Union[str, int, ogr.DataSource, gdal.Dataset,
osr.SpatialReference],
out_path: Optional[str] = None,
copy_if_same: bool = False,
overwrite: bool = True,
) -> Union[List[str], str]:
"""Reprojects a vector given a target projection.
Args:
vector (path | vector): The vector to reproject.
projection (str | int | vector | raster): The projection is infered from
the input. The input can be: WKT proj, EPSG proj, Proj, or read from a
vector or raster datasource either from path or in-memory.
**kwargs:
out_path (path | None): The destination to save to. If None then
the output is an in-memory raster.
overwite (bool): Is it possible to overwrite the out_path if it exists.
Returns:
An in-memory vector. If an out_path is given, the output is a string containing
the path to the newly created vecotr.
"""
type_check(vector, [str, ogr.DataSource], "vector")
type_check(
projection,
[str, int, ogr.DataSource, gdal.Dataset, osr.SpatialReference],
"projection",
)
type_check(out_path, [str], "out_path", allow_none=True)
type_check(copy_if_same, [bool], "copy_if_same")
type_check(overwrite, [bool], "overwrite")
vector_list, path_list = ready_io_vector(vector,
out_path,
overwrite=overwrite)
output = []
for index, in_vector in enumerate(vector_list):
output.append(
internal_reproject_vector(
in_vector,
projection,
out_path=path_list[index],
copy_if_same=copy_if_same,
overwrite=overwrite,
))
if isinstance(vector, list):
return output
return output[0]
def rasters_intersect(
raster1: Union[list, str, gdal.Dataset],
raster2: Union[list, str, gdal.Dataset],
) -> bool:
"""Checks if two rasters intersect."""
type_check(raster1, [list, str, gdal.Dataset], "raster1")
type_check(raster2, [list, str, gdal.Dataset], "raster2")
meta1 = raster_to_metadata(raster1, create_geometry=True)
meta2 = raster_to_metadata(raster2, create_geometry=True)
return meta1["extent_geom_latlng"].Intersects(meta2["extent_geom_latlng"])
def raster_to_disk(
raster: Union[List[Union[str, gdal.Dataset]], str, gdal.Dataset],
out_path: Union[List[str], str],
overwrite: bool = True,
creation_options: Union[list, None] = None,
) -> Union[List[str], str]:
"""Saves or copies a raster to disk. Can be used to change datatype.
Input is either a filepath to a raster or a GDAL.Dataset.
The driver is infered from the file extension.
Args:
raster (path | Dataset): The raster to save to disk.
out_path (path): The destination to save to.
**kwargs:
overwrite (bool): If the file exists, should it be overwritten?
creation_options (list): GDAL creation options. Defaults are:
"TILED=YES"
"NUM_THREADS=ALL_CPUS"
"BIGG_TIF=YES"
"COMPRESS=LZW"
opened (bool): Should the resulting raster be opened
or a path.
Returns:
The filepath for the newly created raster.
"""
type_check(raster, [list, str, gdal.Dataset], "raster")
type_check(out_path, [list, str], "out_path")
type_check(overwrite, [bool], "overwrite")
type_check(creation_options, [list], "creation_options", allow_none=True)
if creation_options is None:
creation_options = []
if not os.path.dirname(os.path.abspath(out_path)):
raise ValueError(
f"Output folder does not exist. Please create first. {out_path}")
raster_list, path_list = ready_io_raster(raster, out_path, overwrite)
output: List[str] = []
for index, in_raster in enumerate(raster_list):
path = _raster_to_disk(
in_raster,
path_list[index],
overwrite=overwrite,
creation_options=default_options(creation_options),
)
output.append(path)
if isinstance(raster, list):
return output
return output[0]
def internal_singlepart_to_multipart(
vector: Union[str, ogr.DataSource],
out_path: Optional[str] = None,
overwrite: bool = True,
add_index: bool = True,
process_layer: int = -1,
) -> str:
type_check(vector, [str, ogr.DataSource], "vector")
type_check(out_path, [str], "out_path", allow_none=True)
type_check(overwrite, [bool], "overwrite")
type_check(add_index, [bool], "add_index")
type_check(process_layer, [int], "process_layer")
vector_list, path_list = ready_io_vector(vector, out_path, overwrite=overwrite)
ref = open_vector(vector_list[0])
out_name = path_list[0]
out_format = path_to_driver_vector(out_name)
driver = ogr.GetDriverByName(out_format)
overwrite_required(out_name, overwrite)
metadata = internal_vector_to_metadata(ref)
remove_if_overwrite(out_name, overwrite)
destination: ogr.DataSource = driver.CreateDataSource(out_name)
for index, layer_meta in enumerate(metadata["layers"]):
if process_layer != -1 and index != process_layer:
continue
name = layer_meta["layer_name"]
geom = layer_meta["column_geom"]
sql = f"SELECT ST_Collect({geom}) AS geom FROM {name};"
result = ref.ExecuteSQL(sql, dialect="SQLITE")
destination.CopyLayer(result, name, ["OVERWRITE=YES"])
if add_index:
vector_add_index(destination)
destination.FlushCache()
return out_name
def raster_set_datatype(
raster: Union[List[Union[str, gdal.Dataset]], str, gdal.Dataset],
dtype: str,
out_path: Optional[Union[List[str], str]],
overwrite: bool = True,
creation_options: Union[list, None] = None,
) -> Union[List[str], str]:
"""Changes the datatype of a raster.
Args:
raster (path | Dataset): The raster(s) to convert.
dtype (str): The destination datatype: Can be float32, uint8 etc..
**kwargs:
out_path (str | None): The destination of the output. If none,
a memory raster with a random name is generated.
creation_options (list): A list of options for the GDAL creation. Only
used if an outpath is specified. Defaults are:
"TILED=YES"
"NUM_THREADS=ALL_CPUS"
"BIGG_TIF=YES"
"COMPRESS=LZW"
Returns:
A path to the newly created raster.
"""
type_check(raster, [list, str, gdal.Dataset], "raster")
type_check(dtype, [str], "dtype")
type_check(out_path, [list, str], "out_path", allow_none=True)
type_check(creation_options, [list], "creation_options", allow_none=True)
raster_list, path_list = ready_io_raster(raster, out_path, overwrite)
output = []
for index, in_raster in enumerate(raster_list):
path = _raster_set_datatype(
in_raster,
dtype,
out_path=path_list[index],
overwrite=overwrite,
creation_options=default_options(creation_options),
)
output.append(path)
if isinstance(raster, list):
return output
return output[0]
def shape_to_blockshape(shape: tuple, block_shape: tuple,
offset: tuple) -> list:
"""Calculates the shape of the output array.
Args:
shape (tuple | list): The shape if the original raster. (1980, 1080, 3)
block_shape (tuple | list): The size of the blocks eg. (64, 64)
offset (tuple, list): An initial offset for the array eg. (32, 32)
Returns:
A list with the modified shape.
"""
type_check(shape, [tuple], "shape")
type_check(block_shape, [tuple], "block_shape")
type_check(offset, [tuple], "offset")
# import pdb
# pdb.set_trace()
assert len(offset) == 2, "Offset has to be two dimensional."
assert len(shape) == 3, "Shape has to be three dimensional."
assert len(block_shape) == 2, "Shape of block has to be two dimensional."
base_shape = list(shape)
for index, value in enumerate(offset):
base_shape[index] = base_shape[index] - value
sizes = [base_shape[0] // block_shape[0], base_shape[1] // block_shape[1]]
return sizes
def open_raster(
raster: Union[list[str, gdal.Dataset], str, gdal.Dataset],
writeable: bool = True,
) -> Union[list[gdal.Dataset], gdal.Dataset]:
"""Opens a raster to a gdal.Dataset class.
Args:
raster (list | path | Dataset): A path to a raster or a GDAL dataframe.
convert_mem_driver (bool): Converts MEM driver rasters to /vsimem/ Gtiffs.
writable (bool): Should the opened raster be writeable.
Returns:
A gdal.Dataset
"""
type_check(raster, [list, str, gdal.Dataset], "raster")
type_check(writeable, [bool], "writeable")
return_list = []
if isinstance(raster, list):
input_list = raster
else:
input_list = [raster]
for readied_raster in input_list:
if isinstance(readied_raster, str):
if path_is_in_memory(readied_raster) or file_exists(
readied_raster):
return_list.append(_open_raster(readied_raster, writeable))
else:
raise ValueError(f"Path does not exists: {readied_raster}")
elif isinstance(readied_raster, gdal.Dataset):
return_list.append(readied_raster)
if isinstance(raster, list):
return return_list
return return_list[0]
def raster_to_memory(
raster: Union[List[Union[str, gdal.Dataset]], str, gdal.Dataset],
memory_path: Optional[Union[List[Union[str, gdal.Dataset]], str]] = None,
) -> Union[List[str], str]:
"""Takes a file path or a gdal raster dataset and copies
it to memory.
Args:
file_path (path | Dataset): A path to a raster or a GDAL dataframe.
**kwargs:
memory_path (str | None): If a path is provided, uses the
appropriate driver and uses the VSIMEM gdal system.
Example: raster_to_memory(clip_ref, "clip_geom.gpkg")
/vsimem/ is autumatically added.
Returns:
A gdal.Dataset copied into memory.
"""
type_check(raster, [list, str, gdal.Dataset], "raster")
type_check(memory_path, [list, str], "memory_path", allow_none=True)
raster_list, out_paths = ready_io_raster(raster,
out_path=memory_path,
overwrite=True)
results: List[str] = []
for index, in_raster in enumerate(raster_list):
result = _raster_to_memory(in_raster, memory_path=out_paths[index])
if not isinstance(result, dict):
raise Exception(f"Error while parsing metadata for: {in_raster}")
results.append(result)
if not isinstance(raster, list):
return results[0]
return results
def _raster_set_datatype(
raster: Union[str, gdal.Dataset],
dtype: str,
out_path: Optional[str],
overwrite: bool = True,
creation_options: Union[list, None] = None,
) -> str:
"""OBS: INTERNAL: Single output.
Changes the datatype of a raster.
"""
type_check(raster, [str, gdal.Dataset], "raster")
type_check(dtype, [str], "dtype")
type_check(out_path, [list, str], "out_path", allow_none=True)
type_check(creation_options, [list], "creation_options", allow_none=True)
ref = open_raster(raster)
metadata = raster_to_metadata(ref)
path = out_path
if path is None:
name = metadata["name"]
path = f"/vsimem/{name}_{uuid4().int}.tif"
driver = gdal.GetDriverByName(path_to_driver_raster(path))
remove_if_overwrite(path, overwrite)
copy = driver.Create(
path,
metadata["height"],
metadata["width"],
metadata["band_count"],
translate_datatypes(dtype),
default_options(creation_options),
)
copy.SetProjection(metadata["projection"])
copy.SetGeoTransform(metadata["transform"])
array = raster_to_array(ref)
for band_idx in range(metadata["band_count"]):
band = copy.GetRasterBand(band_idx + 1)
band.WriteArray(array[:, :, band_idx])
band.SetNoDataValue(metadata["nodata_value"])
return path
def raster_has_nodata_value(
raster: Union[List[Union[gdal.Dataset, str]], gdal.Dataset, str],
) -> Union[bool, List[bool]]:
"""Check if a raster or a list of rasters contain nodata values
Args:
raster (path | raster | list): The raster(s) to check for nodata values.
Returns:
True if input raster has nodata values. If a list is the input, the output
is a list of booleans indicating if the input raster has nodata values.
"""
type_check(raster, [list, str, gdal.Dataset], "raster")
nodata_values = []
rasters = get_raster_path(raster, return_list=True)
for internal_raster in rasters:
if not is_raster(internal_raster):
raise ValueError(f"Input raster is invalid: {internal_raster}")
raster_metadata = raster_to_metadata(internal_raster)
if not isinstance(raster_metadata, dict):
raise Exception("Metadata is in the wrong format.")
raster_nodata = raster_metadata["nodata_value"]
if raster_nodata is not None:
nodata_values.append(True)
else:
nodata_values.append(False)
if isinstance(raster, list):
return nodata_values
else:
return nodata_values[0]
def vector_add_shapes(
vector: Union[List[Union[str, ogr.DataSource]], str, ogr.DataSource],
shapes: list = [
"area", "perimeter", "ipq", "hull", "compactness", "centroid"
],
) -> Union[List[str], str]:
"""Adds shape calculations to a vector such as area and perimeter.
Can also add compactness measurements.
Args:
vector (path | vector): The vector to add shapes to.
**kwargs:
shapes (list): The shapes to calculate. The following a possible:
* Area (In same unit as projection)
* Perimeter (In same unit as projection)
* IPQ (0-1) given as (4*Pi*Area)/(Perimeter ** 2)
* Hull Area (The area of the convex hull. Same unit as projection)
* Compactness (0-1) given as sqrt((area / hull_area) * ipq)
* Centroid (Coordinate of X and Y)
Returns:
Either the path to the updated vector or a list of the input vectors.
"""
type_check(vector, [list, str, ogr.DataSource], "vector")
type_check(shapes, [list], "shapes")
vector_list = to_vector_list(vector)
output = []
for in_vector in vector_list:
output.append(internal_vector_add_shapes(in_vector, shapes=shapes))
if isinstance(vector, list):
return output
return output[0]
def vector_get_fids(
vector: Union[str, ogr.DataSource], process_layer: int = 0
) -> np.ndarray:
type_check(vector, [str, ogr.DataSource], "vector")
type_check(process_layer, [int], "process_layer")
metadata = internal_vector_to_metadata(vector)
features = metadata["layers"][0]["feature_count"]
ref = open_vector(vector)
layer = ref.GetLayer(process_layer)
if layer is None:
raise Exception(f"Requested a non-existing layer: layer_idx={process_layer}")
fid_list = np.empty(features, dtype=int)
for index in range(features):
feature = layer.GetNextFeature()
fid_list[index] = feature.GetFID()
layer.ResetReading()
return fid_list
def vector_to_memory(
vector: Union[List[Union[str, ogr.DataSource]], str, ogr.DataSource],
memory_path: Optional[Union[List[str], str]] = None,
copy_if_already_in_memory: bool = False,
layer_to_extract: int = -1,
) -> Union[List[str], str]:
"""Copies a vector source to memory.
Args:
vector (list | path | DataSource): The vector to copy to memory
**kwargs:
memory_path (str | None): If a path is provided, uses the
appropriate driver and uses the VSIMEM gdal system.
Example: vector_to_memory(clip_ref.tif, "clip_geom.gpkg")
/vsimem/ is autumatically added.
layer_to_extract (int | None): The layer in the vector to copy.
if None is specified, all layers are copied.
opened (bool): If a memory path is specified, the default is
to return a path. If open is supplied. The vector is opened
before returning.
Returns:
An in-memory ogr.DataSource. If a memory path was provided a
string for the in-memory location is returned.
"""
type_check(vector, [list, str, ogr.DataSource], "vector")
type_check(memory_path, [list, str],
"memory_pathout_path",
allow_none=True)
type_check(layer_to_extract, [int], "layer_to_extract")
vector_list, path_list = ready_io_vector(vector, memory_path)
output = []
for index, in_vector in enumerate(vector_list):
path = path_list[index]
output.append(
internal_vector_to_memory(
in_vector,
memory_path=path,
layer_to_extract=layer_to_extract,
copy_if_already_in_memory=copy_if_already_in_memory,
))
if isinstance(vector, list):
return output
return output[0]
def internal_vector_to_memory(
vector: Union[str, ogr.DataSource],
memory_path: Optional[str] = None,
copy_if_already_in_memory: bool = True,
layer_to_extract: int = -1,
) -> str:
"""OBS: Internal. Single output.
Copies a vector source to memory.
"""
type_check(vector, [str, ogr.DataSource], "vector")
type_check(memory_path, [str], "memory_path", allow_none=True)
type_check(layer_to_extract, [int], "layer_to_extract")
ref = open_vector(vector)
path = get_vector_path(ref)
metadata = internal_vector_to_metadata(ref)
name = metadata["name"]
if not copy_if_already_in_memory and metadata["in_memory"]:
if layer_to_extract == -1:
return path
if memory_path is not None:
if memory_path[0:8] == "/vsimem/":
vector_name = memory_path
else:
vector_name = f"/vsimem/{memory_path}"
driver = ogr.GetDriverByName(path_to_driver_vector(memory_path))
else:
vector_name = f"/vsimem/{name}_{uuid4().int}.gpkg"
driver = ogr.GetDriverByName("GPKG")
if driver is None:
raise Exception(f"Error while parsing driver for: {vector}")
copy = driver.CreateDataSource(vector_name)
for layer_idx in range(metadata["layer_count"]):
if layer_to_extract is not None and layer_idx != layer_to_extract:
continue
layername = metadata["layers"][layer_idx]["layer_name"]
copy.CopyLayer(ref.GetLayer(layer_idx), layername, ["OVERWRITE=YES"])
return vector_name
def vector_get_attribute_table(
vector: Union[str, ogr.DataSource],
process_layer: int = 0,
include_geom: bool = False,
) -> pd.DataFrame:
type_check(vector, [str, ogr.DataSource], "vector")
type_check(process_layer, [int], "process_layer")
type_check(include_geom, [bool], "include_geom")
ref = open_vector(vector)
metadata = internal_vector_to_metadata(
ref, process_layer=process_layer, create_geometry=False
)
attribute_table_header = None
feature_count = None
attribute_table_header = metadata["layers"][process_layer]["field_names"]
feature_count = metadata["layers"][process_layer]["feature_count"]
attribute_table = []
layer = ref.GetLayer(process_layer)
for _ in range(feature_count):
feature = layer.GetNextFeature()
attributes = [feature.GetFID()]
for field_name in attribute_table_header:
attributes.append(feature.GetField(field_name))
if include_geom:
geom_defn = feature.GetGeometryRef()
attributes.append(geom_defn.ExportToIsoWkt())
attribute_table.append(attributes)
attribute_table_header.insert(0, "fid")
if include_geom:
attribute_table_header.append("geom")
df = pd.DataFrame(attribute_table, columns=attribute_table_header)
return df
def internal_vector_to_disk(
vector: Union[str, ogr.DataSource],
out_path: str,
overwrite: bool = True,
) -> str:
"""OBS: Internal. Single output.
Copies a vector source to disk.
"""
type_check(vector, [str, ogr.DataSource], "vector")
type_check(out_path, [str], "out_path")
type_check(overwrite, [bool], "overwrite")
overwrite_required(out_path, overwrite)
datasource = open_vector(vector)
metadata = internal_vector_to_metadata(vector)
if not os.path.dirname(os.path.abspath(out_path)):
raise ValueError(
f"Output folder does not exist. Please create first. {out_path}")
driver = ogr.GetDriverByName(path_to_driver_vector(out_path))
if driver is None:
raise Exception(f"Error while parsing driver for: {vector}")
remove_if_overwrite(out_path, overwrite)
copy = driver.CreateDataSource(out_path)
for layer_idx in range(metadata["layer_count"]):
layer_name = metadata["layers"][layer_idx]["layer_name"]
copy.CopyLayer(datasource.GetLayer(layer_idx), str(layer_name),
["OVERWRITE=YES"])
# Flush to disk
copy = None
return out_path
def vector_to_metadata(
vector: Union[List[Union[str, ogr.DataSource]], Union[ogr.DataSource,
str]],
process_layer: int = -1,
create_geometry: bool = True,
) -> Union[List[Metadata_vector], Metadata_vector]:
"""Creates a dictionary with metadata about the vector layer.
Args:
vector (path | DataSource): The vector to analyse.
**kwargs:
create_geometry (bool): Should the metadata include a
footprint of the raster in wgs84. Requires a reprojection
check do not use it if not required and performance is important.
Returns:
A dictionary containing the metadata.
"""
type_check(vector, [list, str, ogr.DataSource], "vector")
type_check(process_layer, [int], "process_layer")
type_check(create_geometry, [bool], "create_geometry")
vector_list = to_vector_list(vector)
output: List[Metadata_vector] = []
for in_vector in vector_list:
output.append(
internal_vector_to_metadata(in_vector,
process_layer=process_layer,
create_geometry=create_geometry))
if isinstance(vector, list):
return output
return output[0]
def merge_vectors(
vectors: List[Union[str, ogr.DataSource]],
out_path: Optional[str] = None,
preserve_fid: bool = True,
) -> str:
"""Merge vectors to a single geopackage."""
type_check(vectors, [list], "vector")
type_check(out_path, [str], "out_path", allow_none=True)
type_check(preserve_fid, [bool], "preserve_fid")
vector_list = to_vector_list(vectors)
out_driver = "GPKG"
out_format = ".gpkg"
out_target = f"/vsimem/clipped_{uuid4().int}{out_format}"
if out_path is not None:
out_target = out_path
out_driver = path_to_driver_vector(out_path)
out_format = path_to_ext(out_path)
driver = ogr.GetDriverByName(out_driver)
merged_ds: ogr.DataSource = driver.CreateDataSource(out_target)
for vector in vector_list:
ref = open_vector(vector)
metadata = internal_vector_to_metadata(ref)
for layer in metadata["layers"]:
name = layer["layer_name"]
merged_ds.CopyLayer(ref.GetLayer(name), name, ["OVERWRITE=YES"])
merged_ds.FlushCache()
return out_target
def vector_to_disk(
vector: Union[List[Union[str, ogr.DataSource]], str, ogr.DataSource],
out_path: Union[List[str], str],
overwrite: bool = True,
) -> Union[List[str], str]:
"""Copies a vector source to disk.
Args:
vector (path | DataSource): The vector to copy to disk
out_path (path): The destination to save to.
**kwargs:
overwite (bool): Is it possible to overwrite the out_path if it exists.
Returns:
An path to the created vector.
"""
type_check(vector, [str, ogr.DataSource], "vector")
type_check(out_path, [str], "out_path")
type_check(overwrite, [bool], "overwrite")
vector_list, path_list = ready_io_vector(vector,
out_path,
overwrite=overwrite)
output = []
for index, in_vector in enumerate(vector_list):
path = path_list[index]
output.append(
internal_vector_to_disk(in_vector, path, overwrite=overwrite))
if isinstance(vector, list):
return output
return output[0]
def intersect_vector(
vector: Union[List[Union[str, ogr.DataSource]], str, ogr.DataSource],
clip_geom: Union[
List[Union[str, ogr.DataSource, gdal.Dataset]],
gdal.Dataset,
ogr.DataSource,
str,
],
out_path: str = None,
to_extent: bool = False,
process_layer: int = 0,
process_layer_clip: int = 0,
add_index: bool = True,
preserve_fid: bool = True,
overwrite: bool = True,
) -> Union[List[str], str]:
"""Clips a vector to a geometry."""
type_check(vector, [ogr.DataSource, str, list], "vector")
type_check(clip_geom, [ogr.DataSource, gdal.Dataset, str, list, tuple], "clip_geom")
type_check(out_path, [str], "out_path", allow_none=True)
type_check(to_extent, [bool], "to_extent")
type_check(process_layer, [int], "process_layer")
type_check(process_layer_clip, [int], "process_layer_clip")
type_check(add_index, [bool], "add_index")
type_check(preserve_fid, [bool], "preserve_fid")
type_check(overwrite, [bool], "overwrite")
vector_list, path_list = ready_io_vector(vector, out_path, overwrite=overwrite)
output = []
for index, in_vector in enumerate(vector_list):
output.append(
internal_intersect_vector(
in_vector,
clip_geom,
out_path=path_list[index],
to_extent=to_extent,
process_layer=process_layer,
process_layer_clip=process_layer_clip,
add_index=add_index,
preserve_fid=preserve_fid,
overwrite=overwrite,
)
)
if isinstance(vector, list):
return output
return output[0]
def internal_intersect_vector(
vector: Union[str, ogr.DataSource],
clip_geom: Union[str, ogr.DataSource, gdal.Dataset],
out_path: Optional[str] = None,
to_extent: bool = False,
process_layer: int = 0,
process_layer_clip: int = 0,
add_index: bool = True,
preserve_fid: bool = True,
overwrite: bool = True,
return_bool: bool = False,
) -> str:
"""Clips a vector to a geometry.
Returns:
A clipped ogr.Datasource or the path to one.
"""
type_check(vector, [ogr.DataSource, str, list], "vector")
type_check(clip_geom, [ogr.DataSource, gdal.Dataset, str, list, tuple], "clip_geom")
type_check(out_path, [str], "out_path", allow_none=True)
type_check(to_extent, [bool], "to_extent")
type_check(process_layer, [int], "process_layer")
type_check(process_layer_clip, [int], "process_layer_clip")
type_check(add_index, [bool], "add_index")
type_check(preserve_fid, [bool], "preserve_fid")
type_check(overwrite, [bool], "overwrite")
_vector_list, path_list = ready_io_vector(vector, out_path, overwrite=overwrite)
out_name = path_list[0]
match_projection = internal_reproject_vector(clip_geom, vector)
geometry_to_clip = open_vector(match_projection)
merged = open_vector(merge_vectors([vector, match_projection]))
if add_index:
vector_add_index(merged)
vector_metadata = internal_vector_to_metadata(vector)
vector_layername = vector_metadata["layers"][process_layer]["layer_name"]
vector_geom_col = vector_metadata["layers"][process_layer]["column_geom"]
clip_geom_metadata = internal_vector_to_metadata(geometry_to_clip)
clip_geom_layername = clip_geom_metadata["layers"][process_layer_clip]["layer_name"]
clip_geom_col = clip_geom_metadata["layers"][process_layer_clip]["column_geom"]
if return_bool:
sql = f"SELECT A.* FROM '{vector_layername}' A, '{clip_geom_layername}' B WHERE ST_INTERSECTS(A.{vector_geom_col}, B.{clip_geom_col});"
else:
sql = f"SELECT A.* FROM '{vector_layername}' A, '{clip_geom_layername}' B WHERE ST_INTERSECTS(A.{vector_geom_col}, B.{clip_geom_col});"
result = merged.ExecuteSQL(sql, dialect="SQLITE")
if return_bool:
if result.GetFeatureCount() == 0:
return False
else:
return True
driver = ogr.GetDriverByName(path_to_driver_vector(out_name))
destination: ogr.DataSource = driver.CreateDataSource(out_name)
destination.CopyLayer(result, vector_layername, ["OVERWRITE=YES"])
if destination is None:
raise Exception("Error while running intersect.")
destination.FlushCache()
return out_name
def internal_vector_add_shapes(
vector: Union[str, ogr.DataSource],
shapes: list = [
"area", "perimeter", "ipq", "hull", "compactness", "centroid"
],
) -> str:
"""OBS: Internal. Single output.
Adds shape calculations to a vector such as area and perimeter.
Can also add compactness measurements.
"""
type_check(vector, [str, ogr.DataSource], "vector")
type_check(shapes, [list], "shapes")
datasource = open_vector(vector)
out_path = get_vector_path(datasource)
metadata = internal_vector_to_metadata(datasource)
for index in range(metadata["layer_count"]):
vector_current_fields = metadata["layers"][index]["field_names"]
vector_layer = datasource.GetLayer(index)
vector_layer.StartTransaction()
# Add missing fields
for attribute in shapes:
if attribute == "centroid":
if "centroid_x" not in vector_current_fields:
field_defn = ogr.FieldDefn("centroid_x", ogr.OFTReal)
vector_layer.CreateField(field_defn)
if "centroid_y" not in vector_current_fields:
field_defn = ogr.FieldDefn("centroid_y", ogr.OFTReal)
vector_layer.CreateField(field_defn)
elif attribute not in vector_current_fields:
field_defn = ogr.FieldDefn(attribute, ogr.OFTReal)
vector_layer.CreateField(field_defn)
vector_feature_count = vector_layer.GetFeatureCount()
for i in range(vector_feature_count):
vector_feature = vector_layer.GetNextFeature()
try:
vector_geom = vector_feature.GetGeometryRef()
except:
vector_geom.Buffer(0)
Warning("Invalid geometry at : ", i)
if vector_geom is None:
raise Exception("Invalid geometry. Could not fix.")
centroid = vector_geom.Centroid()
vector_area = vector_geom.GetArea()
vector_perimeter = vector_geom.Boundary().Length()
if "ipq" or "compact" in shapes:
vector_ipq = 0
if vector_perimeter != 0:
vector_ipq = (4 * np.pi *
vector_area) / vector_perimeter**2
if "centroid" in shapes:
vector_feature.SetField("centroid_x", centroid.GetX())
vector_feature.SetField("centroid_y", centroid.GetY())
if "hull" in shapes or "compact" in shapes:
vector_hull = vector_geom.ConvexHull()
hull_area = vector_hull.GetArea()
hull_peri = vector_hull.Boundary().Length()
hull_ratio = float(vector_area) / float(hull_area)
compactness = np.sqrt(float(hull_ratio) * float(vector_ipq))
if "area" in shapes:
vector_feature.SetField("area", vector_area)
if "perimeter" in shapes:
vector_feature.SetField("perimeter", vector_perimeter)
if "ipq" in shapes:
vector_feature.SetField("ipq", vector_ipq)
if "hull" in shapes:
vector_feature.SetField("hull_area", hull_area)
vector_feature.SetField("hull_peri", hull_peri)
vector_feature.SetField("hull_ratio", hull_ratio)
if "compact" in shapes:
vector_feature.SetField("compact", compactness)
vector_layer.SetFeature(vector_feature)
progress(i, vector_feature_count, name="shape")
vector_layer.CommitTransaction()
return out_path
def ready_io_vector(
vector: Union[List[Union[str, ogr.DataSource]], str, ogr.DataSource],
out_path: Optional[Union[List[str], str]],
overwrite: bool = True,
add_uuid: bool = False,
prefix: str = "",
postfix: str = "",
) -> Tuple[List[str], List[str]]:
type_check(vector, [list, str, ogr.DataSource], "vector")
type_check(out_path, [list, str], "out_path", allow_none=True)
type_check(overwrite, [bool], "overwrite")
type_check(prefix, [str], "prefix")
type_check(postfix, [str], "postfix")
vector_list = to_vector_list(vector)
if isinstance(out_path, list):
if len(vector_list) != len(out_path):
raise ValueError(
"The length of vector_list must equal the length of the out_path"
)
# Check if folder exists and is required.
if len(vector_list) > 1 and isinstance(out_path, str):
if not os.path.dirname(os.path.abspath(out_path)):
raise ValueError(
f"Output folder does not exist. Please create first. {out_path}"
)
# Generate output names
path_list: List[str] = []
for index, in_vector in enumerate(vector_list):
metadata = internal_vector_to_metadata(in_vector)
name = metadata["name"]
if add_uuid:
uuid = uuid4().int
else:
uuid = ""
if out_path is None:
path = f"/vsimem/{prefix}{name}{uuid}{postfix}.gpkg"
elif isinstance(out_path, str):
if folder_exists(out_path):
path = os.path.join(out_path,
f"{prefix}{name}{uuid}{postfix}.tif")
else:
path = out_path
elif isinstance(out_path, list):
if out_path[index] is None:
path = f"/vsimem/{prefix}{name}{uuid}{postfix}.tif"
elif isinstance(out_path[index], str):
path = out_path[index]
else:
raise ValueError(f"Unable to parse out_path: {out_path}")
else:
raise ValueError(f"Unable to parse out_path: {out_path}")
overwrite_required(path, overwrite)
path_list.append(path)
return (vector_list, path_list)