Source code for rios.inputcollection
"""
This module contains the InputCollection and InputIterator
classes. These classes are for ImageReader to keep track
of the inputs it has and deal with resampling.
"""
# This file is part of RIOS - Raster I/O Simplification
# Copyright (C) 2012 Sam Gillingham, Neil Flood
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
import os
import sys
import tempfile
from . import imageio
from . import rioserrors
from . import pixelgrid
from osgeo import gdal
[docs]class InputIterator(object):
"""
Class to allow iteration across an InputCollection instance.
Do not instantiate this class directly - it is created
by InputCollection.__iter__().
See http://docs.python.org/library/stdtypes.html#typeiter
for a description of how this works. There is another way,
see: http://docs.python.org/reference/expressions.html#yieldexpr
but it seemed too much like Windows 3.1 programming which
scared me!
Returns a image name, GDAL Dataset, PixelGridDefn, nullvallist and datatype
for each iteration
"""
def __init__(self, collection):
# collection = an InputCollection instance
self.collection = collection
self.index = 0 # start at first one
def __iter__(self):
# For iteration support - just return self.
return self
def __next__(self):
# for iteration support. Raises a StopIteration
# if we have read beyond the end of the collection
if self.index >= len(self.collection.datasetList):
raise StopIteration()
# pull out the values we want to return
image = self.collection.imageList[self.index]
ds = self.collection.datasetList[self.index]
pixgrid = self.collection.pixgridList[self.index]
nullvals = self.collection.nullValList[self.index]
datatype = self.collection.dataTypeList[self.index]
# so next time we are looking at the next one
self.index += 1
return (image, ds, pixgrid, nullvals, datatype)
[docs]class InputCollection(object):
"""
InputCollection class. Keeps all of the inputs and the
information we need about them in one place.
Gets passed a list of filenames and opens them, creates
a PixelGridDefn and pulls out their null values.
Setting the reference dataset is done via setReference()
(is first dataset by default). Resampling can be
performed by resampleToReference().
Use checkAllMatch() to see if resampling is necessary.
"""
def __init__(self, imageList, loggingstream=sys.stdout):
"""
Constructor. imageList is a list of file names that
need to be opened. An ImageOpenException() is raised
should opening fail.
Set loggingstream to a file like object if you wish
logging about resampling to be sent somewhere else
rather than stdout.
"""
self.loggingstream = loggingstream
# initialise our lists
self.imageList = []
self.datasetList = []
self.pixgridList = []
self.nullValList = [] # NB: a list of lists
self.dataTypeList = [] # numpy types
self.filestoremove = [] # any temp resampled files
# go thru each image
for image in imageList:
# try and open it
ds = gdal.Open(str(image))
if ds is None:
msg = 'Unable to Open %s' % image
raise rioserrors.ImageOpenError(msg)
# create a PixelGridDefn for the dataset
pixGrid = self.makePixGridFromDataset(ds)
# Stash the null values for each band
dsNullValList = []
for i in range(ds.RasterCount):
nullVal = ds.GetRasterBand(i + 1).GetNoDataValue()
dsNullValList.append(nullVal)
# get the datatype of band 1
gdaldatatype = ds.GetRasterBand(1).DataType
numpytype = imageio.GDALTypeToNumpyType(gdaldatatype)
# store the values in our lists
self.imageList.append(image)
self.datasetList.append(ds)
self.pixgridList.append(pixGrid)
self.nullValList.append(dsNullValList)
self.dataTypeList.append(numpytype)
# by default the reference dataset is the first file
self.referencePixGrid = self.pixgridList[0]
def __del__(self):
"""
Remove all resampled files. This does not appear
to get called if there is an exception in the
reader, so there is an exception handler and a finally in
ImageReader.readBlock which calls cleanup directly.
"""
self.cleanup()
[docs] def close(self):
"""
Closes all open datasets
"""
for ds in self.datasetList:
del ds
self.datasetList = []
self.cleanup()
[docs] def cleanup(self):
"""
Removes any temp files. To be called from destructor?
Seems not given Neil's experience - needs to be called
from finally clause - not sure how to enforce this
in user's script....
"""
for f in self.filestoremove:
if os.path.exists(f):
try:
os.remove(f)
except rioserrors.PermissionError:
# ignore any 'file in use' errors on Windows
# This is only a problem when useVRT=False as we are dealing
# with an actual dataset that cannot be deleted (rather than a VRT
# which can for some reason). Fixing this properly is going to be
# quite hard (the info object and maybe others still hold the
# dataset open) so let's ignore for now.
pass
self.filestoremove = []
def __len__(self):
# see http://docs.python.org/reference/datamodel.html#emulating-container-types
return len(self.datasetList)
def __getitem__(self, key):
# see http://docs.python.org/reference/datamodel.html#emulating-container-types
# for indexing, returns:
# imagename, dataset, pixgrid,nullValList, datatype
# if key outside range, then the individual lists
# should raise KeyError
image = self.imageList[key]
ds = self.datasetList[key]
pixgrid = self.pixgridList[key]
nullValList = self.nullValList[key]
datatype = self.dataTypeList[key]
return (image, ds, pixgrid, nullValList, datatype)
def __iter__(self):
# see http://docs.python.org/reference/datamodel.html#emulating-container-types
return InputIterator(self)
[docs] def setReference(self, refpath=None, refgeotrans=None,
refproj=None, refNCols=None, refNRows=None, refPixgrid=None):
"""
Sets the reference dataset for resampling purposes.
Either set refpath to a path to a GDAL file and all
necessary values will be extracted. Or pass refgeotrans,
refproj, refNCols and refNRows.
"""
# if we have a path, open it with GDAL
if refpath is not None:
ds = gdal.Open(refpath)
if ds is None:
msg = 'Unable to Open %s' % refpath
raise rioserrors.ImageOpenError(msg)
# make a pixgrid from the dataset
self.referencePixGrid = self.makePixGridFromDataset(ds)
# close the dataset
del ds
# else, check we have all the aother params
elif refgeotrans is not None and refproj is not None and refNCols is not None and refNRows is not None:
# create a pixgrid from the info we have
proj = self.specialProjFixes(refproj)
pixGrid = pixelgrid.PixelGridDefn(geotransform=refgeotrans,
nrows=refNRows, ncols=refNCols, projection=proj)
self.referencePixGrid = pixGrid
elif refPixgrid is not None:
# Or maybe we have been given a complete pixel grid already
refPixgrid.projection = self.specialProjFixes(refPixgrid.projection)
self.referencePixGrid = refPixgrid
else:
msg = 'Must pass either refpath or refPixgrid or all the other params'
raise rioserrors.ParameterError(msg)
[docs] def resampleToReference(self, ds, nullValList, workingRegion, resamplemethod,
tempdir='.', useVRT=False, allowOverviewsGdalwarp=False):
"""
Resamples any inputs that do not match the reference, to the
reference image.
ds is the GDAL dataset that needs to be resampled.
nullValList is the list of null values for that dataset.
workingRegion is a PixelGridDefn
resamplemethod is a string containing a method supported by gdalwarp.
Returns the new (temporary) resampled dataset instance.
Do not call directly, use resampleAllToReference()
"""
if useVRT:
ext = '.vrt'
else:
# get the driver from the input dataset
# so we know the default extension, and type
# for the temporary file.
driver = ds.GetDriver()
drivermeta = driver.GetMetadata()
# temp image file name - based on driver extension
ext = ''
if gdal.DMD_EXTENSION in drivermeta:
ext = '.' + drivermeta[gdal.DMD_EXTENSION]
(fileh, temp_image) = tempfile.mkstemp(ext, dir=tempdir)
os.close(fileh)
overviewLevel = 'NONE'
if allowOverviewsGdalwarp:
overviewLevel = 'AUTO'
if useVRT:
driverName = 'VRT'
else:
driverName = driver.ShortName
nullValues = self.makeWarpNullOptions(nullValList)
warpOptions = gdal.WarpOptions(overviewLevel=overviewLevel,
srcSRS=self.specialProjFixes(ds.GetProjection()),
dstSRS=self.referencePixGrid.projection,
outputBounds=[workingRegion.xMin, workingRegion.yMin,
workingRegion.xMax, workingRegion.yMax],
xRes=workingRegion.xRes, yRes=workingRegion.yRes,
format=driverName, resampleAlg=resamplemethod,
srcNodata=nullValues, dstNodata=nullValues)
# delete later - add before the system call as if Ctrl-C is hit
# control does not always return
self.filestoremove.append(temp_image)
usingExceptions = gdal.GetUseExceptions()
gdal.UseExceptions()
try:
newds = gdal.Warp(temp_image, ds, options=warpOptions)
except Exception as e:
msg = ("Error while running gdal.Warp(): {}. Try setting the " +
"RIOS_NO_VRT_FOR_RESAMPLING environment variable " +
"to '1'").format(str(e))
raise rioserrors.GdalWarpError(msg)
finally:
if not usingExceptions:
gdal.DontUseExceptions()
# return the new dataset
return newds
[docs] def resampleAllToReference(self, footprint, resamplemethodlist, tempdir='.', useVRT=False,
allowOverviewsGdalwarp=False):
"""
Reamples all datasets that don't match the reference to the
same as the reference.
footprint is imageio.INTERSECTION, imageio.UNION or imageio.BOUNDS_FROM_REFERENCE
resamplemethod is a string containing a method supported by gdalwarp.
"""
# find the working region based on the footprint
workingRegion = self.findWorkingRegion(footprint)
# go thru each input.
for count in range(len(self.datasetList)):
pixGrid = self.pixgridList[count]
# check the pixgrid of the dataset matches the
# reference
allEqual = True
if not self.referencePixGrid.equalPixSize(pixGrid):
self.loggingstream.write("Pixel sizes don't match %.20f %.20f %.20f %.20f\n" %
(self.referencePixGrid.xRes, pixGrid.xRes, self.referencePixGrid.yRes, pixGrid.yRes))
allEqual = False
elif not self.referencePixGrid.equalProjection(pixGrid):
self.loggingstream.write("Coordinate systems don't match %s %s\n" %
(self.referencePixGrid.projection, pixGrid.projection))
allEqual = False
elif not self.referencePixGrid.alignedWith(pixGrid):
self.loggingstream.write("Images aren't on the same grid\n")
allEqual = False
# did it fail to match completely?
if not allEqual:
# resample the dataset
ds = self.datasetList[count]
nullVals = self.nullValList[count]
resamplemethod = resamplemethodlist[count]
newds = self.resampleToReference(ds, nullVals, workingRegion, resamplemethod,
tempdir, useVRT, allowOverviewsGdalwarp)
# stash the new temp dataset as our input item
self.datasetList[count] = newds
# just assume it has been successfully
# resampled to exactly the workingRegion
self.pixgridList[count] = workingRegion
# close the original dataset - just using temp
# resampled one from now.
del ds
[docs] def checkAllMatch(self):
"""
Returns whether any resampling necessary to match
reference dataset.
Use as a check if no resampling is done that we
can proceed ok.
"""
match = True
for pixGrid in self.pixgridList:
if not self.referencePixGrid.equalPixSize(pixGrid):
match = False
break
elif not self.referencePixGrid.equalProjection(pixGrid):
match = False
break
elif not self.referencePixGrid.alignedWith(pixGrid):
match = False
break
return match
[docs] @staticmethod
def makeWarpNullOptions(nullValList):
"""
Make appropriate options for gdalwarp, to handle
null value properly. If any of the null values
in the list is None, then we can't do it because the null
value is not set on the file, so the return value is None.
Returns a list that can be passed to as srcNodata and
dstNodata params to gdal.Warp().
"""
haveNone = False
for nullVal in nullValList:
if nullVal is None:
haveNone = True
if haveNone:
allNulls = None
else:
nullValStrList = [str(n) for n in nullValList]
allNulls = ' '.join(nullValStrList)
return allNulls
[docs] @staticmethod
def specialProjFixes(projwkt):
"""
Does any special fixes required for the projection. Returns the fixed
projection WKT string.
Specifically this does two things, both of which are to cope with rubbish
that Imagine has put into the projection. Firstly, it removes the
crappy TOWGS84 parameters which Imagine uses for GDA94, and secondly
removes the crappy name which Imagine gives to the correct GDA94.
If neither of these things is found, returns the string unchanged.
"""
dodgyTOWGSstring = "TOWGS84[-16.237,3.51,9.939,1.4157e-06,2.1477e-06,1.3429e-06,1.91e-07]"
properTOWGSstring = "TOWGS84[0,0,0,0,0,0,0]"
if projwkt.find('"GDA94"') > 0 or projwkt.find('"Geocentric_Datum_of_Australia_1994"') > 0:
newWkt = projwkt.replace(dodgyTOWGSstring, properTOWGSstring)
else:
newWkt = projwkt
# Imagine's name for the correct GDA94 also causes problems, so
# replace it with something more standard.
newWkt = newWkt.replace('GDA94-ICSM', 'GDA94')
return newWkt
[docs] @staticmethod
def makePixGridFromDataset(ds):
"""
Make a pixelgrid object from the given dataset.
"""
proj = InputCollection.specialProjFixes(ds.GetProjection())
geotrans = ds.GetGeoTransform()
(nrows, ncols) = (ds.RasterYSize, ds.RasterXSize)
pixGrid = pixelgrid.PixelGridDefn(geotransform=geotrans, nrows=nrows, ncols=ncols,
projection=proj)
return pixGrid
[docs] def findWorkingRegion(self, footprint):
"""
Work out what the combined pixel grid should be, in terms of the
given reference input raster. Returns a PixelGridDefn object.
"""
combinedGrid = pixelgrid.findCommonRegion(self.pixgridList,
self.referencePixGrid, combine=footprint)
return combinedGrid