def read_attribute(fname,
datasetName=None,
standardize=True,
metafile_ext=None):
"""Read attributes of input file into a dictionary
Parameters: fname : str, path/name of data file
datasetName : str, name of dataset of interest, for file with multiple datasets
e.g. unwrapPhase in ifgramStack.h5
coherence in ifgramStack.h5
height in geometryRadar.h5
latitude in geometryRadar.h5
...
standardize : bool, grab standardized metadata key name
Returns: atr : dict, attributes dictionary
"""
fbase, fext = os.path.splitext(os.path.basename(fname))
fext = fext.lower()
if not os.path.isfile(fname):
msg = 'input file not existed: {}\n'.format(fname)
msg += 'current directory: ' + os.getcwd()
raise Exception(msg)
# HDF5 files
if fext in ['.h5', '.he5']:
f = h5py.File(fname, 'r')
g1_list = [i for i in f.keys() if isinstance(f[i], h5py.Group)]
d1_list = [
i for i in f.keys()
if isinstance(f[i], h5py.Dataset) and f[i].ndim >= 2
]
# FILE_TYPE - k
py2_mintpy_stack_files = ['interferograms', 'coherence',
'wrapped'] #obsolete mintpy format
if any(i in d1_list for i in ['unwrapPhase']):
k = 'ifgramStack'
elif any(i in d1_list for i in ['height', 'latitude', 'azimuthCoord']):
k = 'geometry'
elif any(i in g1_list + d1_list
for i in ['timeseries', 'displacement']):
k = 'timeseries'
elif 'HDFEOS' in g1_list:
k = 'HDFEOS'
elif 'recons' in d1_list:
k = 'giantTimeseries'
elif any(i in d1_list for i in ['igram', 'figram']):
k = 'giantIfgramStack'
elif any(i in g1_list for i in py2_mintpy_stack_files):
k = list(set(g1_list) & set(py2_mintpy_stack_files))[0]
elif len(d1_list) > 0:
k = d1_list[0]
elif len(g1_list) > 0:
k = g1_list[0]
else:
raise ValueError('unrecognized file type: ' + fname)
# metadata dict
if k == 'giantTimeseries':
atr = giantTimeseries(fname).get_metadata()
elif k == 'giantIfgramStack':
atr = giantIfgramStack(fname).get_metadata()
else:
if len(f.attrs) > 0 and 'WIDTH' in f.attrs.keys():
atr = dict(f.attrs)
else:
# grab the list of attrs in HDF5 file
global atr_list
def get_hdf5_attrs(name, obj):
global atr_list
if len(obj.attrs) > 0 and 'WIDTH' in obj.attrs.keys():
atr_list.append(dict(obj.attrs))
atr_list = []
f.visititems(get_hdf5_attrs)
# use the attrs with most items
if atr_list:
num_list = [len(i) for i in atr_list]
atr = atr_list[np.argmax(num_list)]
else:
raise ValueError('No attribute WIDTH found in file:',
fname)
# decode string format
for key, value in atr.items():
try:
atr[key] = value.decode('utf8')
except:
atr[key] = value
# attribute identified by MintPy
# 1. FILE_TYPE
atr['FILE_TYPE'] = str(k)
# 2. DATA_TYPE
ds = None
if datasetName and datasetName in f.keys():
ds = f[datasetName]
else:
# get the 1st dataset
global ds_list
def get_hdf5_dataset(name, obj):
global ds_list
if isinstance(obj, h5py.Dataset) and obj.ndim >= 2:
ds_list.append(obj)
ds_list = []
f.visititems(get_hdf5_dataset)
if ds_list:
ds = ds_list[0]
if ds is not None:
atr['DATA_TYPE'] = str(ds.dtype)
f.close()
# 3. PROCESSOR
if 'INSAR_PROCESSOR' in atr.keys():
atr['PROCESSOR'] = atr['INSAR_PROCESSOR']
if 'PROCESSOR' not in atr.keys():
atr['PROCESSOR'] = 'mintpy'
else:
# get metadata file basename
metafile_base = fname
# for SNAP BEAM-DIMAP format
if fname.endswith('.img') and os.path.isfile(
os.path.splitext(fname)[0] + '.hdr'):
metafile_base = os.path.splitext(fname)[0]
# get existing metadata file extensions
metafile_exts = ['.rsc', '.xml', '.aux.xml', '.par', '.hdr']
if metafile_ext:
metafile_exts = [
i for i in metafile_exts if i.endswith(metafile_ext)
]
metafile_exts = [
i for i in metafile_exts if os.path.isfile(metafile_base + i)
]
if len(metafile_exts) == 0:
raise FileNotFoundError(
'No metadata file found for data file: {}'.format(fname))
atr = {}
# PROCESSOR
if any(i.endswith('.hdr')
for i in metafile_exts) and fname.endswith('.img'):
atr['PROCESSOR'] = 'snap'
elif any(i.endswith('.xml') for i in metafile_exts):
atr['PROCESSOR'] = 'isce'
xml_exts = [i for i in metafile_exts if i.endswith('.xml')]
if len(xml_exts) > 0:
atr.update(read_isce_xml(metafile_base + xml_exts[0]))
elif any(i.endswith('.par') for i in metafile_exts):
atr['PROCESSOR'] = 'gamma'
elif any(i.endswith('.rsc') for i in metafile_exts):
if 'PROCESSOR' not in atr.keys():
atr['PROCESSOR'] = 'roipac'
if 'PROCESSOR' not in atr.keys():
atr['PROCESSOR'] = 'mintpy'
# Read metadata file and FILE_TYPE
metafile0 = metafile_base + metafile_exts[0]
while fext in ['.geo', '.rdr', '.full']:
fbase, fext = os.path.splitext(fbase)
if not fext:
fext = fbase
if metafile0.endswith('.rsc'):
atr.update(read_roipac_rsc(metafile0))
if 'FILE_TYPE' not in atr.keys():
atr['FILE_TYPE'] = fext
elif metafile0.endswith('.xml'):
atr.update(read_isce_xml(metafile0))
if 'FILE_TYPE' not in atr.keys():
atr['FILE_TYPE'] = fext #atr.get('image_type', fext)
elif metafile0.endswith('.par'):
atr.update(read_gamma_par(metafile0))
atr['FILE_TYPE'] = fext
elif metafile0.endswith('.hdr'):
atr.update(read_envi_hdr(metafile0))
fbase = os.path.basename(fname).lower()
if fbase.startswith('unw'):
atr['FILE_TYPE'] = '.unw'
elif fbase.startswith(('coh', 'cor')):
atr['FILE_TYPE'] = '.cor'
elif fbase.startswith('phase_ifg'):
atr['FILE_TYPE'] = '.int'
elif 'dem' in fbase:
atr['FILE_TYPE'] = 'dem'
else:
atr['FILE_TYPE'] = atr['file type']
# UNIT
k = atr['FILE_TYPE'].replace('.', '')
if k == 'ifgramStack':
if datasetName and datasetName in datasetUnitDict.keys():
atr['UNIT'] = datasetUnitDict[datasetName]
else:
atr['UNIT'] = 'radian'
elif 'UNIT' not in atr.keys():
if datasetName and datasetName in datasetUnitDict.keys():
atr['UNIT'] = datasetUnitDict[datasetName]
elif k in datasetUnitDict.keys():
atr['UNIT'] = datasetUnitDict[k]
else:
atr['UNIT'] = '1'
# FILE_PATH
atr['FILE_PATH'] = os.path.abspath(fname)
if standardize:
atr = standardize_metadata(atr)
return atr
def read_attribute(fname, datasetName=None, standardize=True, metafile_ext=None):
"""Read attributes of input file into a dictionary
Parameters: fname : str, path/name of data file
datasetName : str, name of dataset of interest, for file with multiple datasets
e.g. unwrapPhase in ifgramStack.h5
coherence in ifgramStack.h5
height in geometryRadar.h5
latitude in geometryRadar.h5
...
standardize : bool, grab standardized metadata key name
Returns: atr : dict, attributes dictionary
"""
fbase, fext = os.path.splitext(os.path.basename(fname))
fext = fext.lower()
if not os.path.isfile(fname):
msg = 'input file not existed: {}\n'.format(fname)
msg += 'current directory: '+os.getcwd()
raise Exception(msg)
# HDF5 files
if fext in ['.h5', '.he5']:
f = h5py.File(fname, 'r')
g1_list = [i for i in f.keys() if isinstance(f[i], h5py.Group)]
d1_list = [i for i in f.keys() if isinstance(f[i], h5py.Dataset) and f[i].ndim >= 2]
# FILE_TYPE - k
py2_mintpy_stack_files = ['interferograms', 'coherence', 'wrapped'] #obsolete mintpy format
if any(i in d1_list for i in ['unwrapPhase']):
k = 'ifgramStack'
elif any(i in d1_list for i in ['height', 'latitude', 'azimuthCoord']):
k = 'geometry'
elif any(i in g1_list+d1_list for i in ['timeseries', 'displacement']):
k = 'timeseries'
elif 'HDFEOS' in g1_list:
k = 'HDFEOS'
elif 'recons' in d1_list:
k = 'giantTimeseries'
elif any(i in d1_list for i in ['igram', 'figram']):
k = 'giantIfgramStack'
elif any(i in g1_list for i in py2_mintpy_stack_files):
k = list(set(g1_list) & set(py2_mintpy_stack_files))[0]
elif len(d1_list) > 0:
k = d1_list[0]
elif len(g1_list) > 0:
k = g1_list[0]
else:
raise ValueError('unrecognized file type: '+fname)
# metadata dict
if k == 'giantTimeseries':
atr = giantTimeseries(fname).get_metadata()
elif k == 'giantIfgramStack':
atr = giantIfgramStack(fname).get_metadata()
else:
if len(f.attrs) > 0 and 'WIDTH' in f.attrs.keys():
atr = dict(f.attrs)
else:
# grab the list of attrs in HDF5 file
global atr_list
def get_hdf5_attrs(name, obj):
global atr_list
if len(obj.attrs) > 0 and 'WIDTH' in obj.attrs.keys():
atr_list.append(dict(obj.attrs))
atr_list = []
f.visititems(get_hdf5_attrs)
# use the attrs with most items
if atr_list:
num_list = [len(i) for i in atr_list]
atr = atr_list[np.argmax(num_list)]
else:
raise ValueError('No attribute WIDTH found in file:', fname)
# decode string format
for key, value in atr.items():
try:
atr[key] = value.decode('utf8')
except:
atr[key] = value
# attribute identified by MintPy
# 1. FILE_TYPE
atr['FILE_TYPE'] = str(k)
# 2. DATA_TYPE
ds = None
if datasetName and datasetName in f.keys():
ds = f[datasetName]
else:
# get the 1st dataset
global ds_list
def get_hdf5_dataset(name, obj):
global ds_list
if isinstance(obj, h5py.Dataset) and obj.ndim >= 2:
ds_list.append(obj)
ds_list = []
f.visititems(get_hdf5_dataset)
if ds_list:
ds = ds_list[0]
if ds is not None:
atr['DATA_TYPE'] = str(ds.dtype)
f.close()
# 3. PROCESSOR
if 'INSAR_PROCESSOR' in atr.keys():
atr['PROCESSOR'] = atr['INSAR_PROCESSOR']
if 'PROCESSOR' not in atr.keys():
atr['PROCESSOR'] = 'mintpy'
else:
# get existing metadata file extensions
metafile_exts = ['.rsc', '.xml', '.aux.xml', '.par', '.hdr']
if metafile_ext:
metafile_exts = [i for i in metafile_exts if i.endswith(metafile_ext)]
metafile_exts = [i for i in metafile_exts if os.path.isfile(fname+i)]
if len(metafile_exts) == 0:
raise FileNotFoundError('No metadata file found for data file: {}'.format(fname))
atr = {}
# PROCESSOR
if any(i.endswith(('.xml', '.hdr')) for i in metafile_exts):
atr['PROCESSOR'] = 'isce'
xml_exts = [i for i in metafile_exts if i.endswith('.xml')]
if len(xml_exts) > 0:
atr.update(read_isce_xml(fname+xml_exts[0]))
elif any(i.endswith('.par') for i in metafile_exts):
atr['PROCESSOR'] = 'gamma'
elif any(i.endswith('.rsc') for i in metafile_exts):
if 'PROCESSOR' not in atr.keys():
atr['PROCESSOR'] = 'roipac'
if 'PROCESSOR' not in atr.keys():
atr['PROCESSOR'] = 'mintpy'
# Read metadata file and FILE_TYPE
metafile0 = fname + metafile_exts[0]
while fext in ['.geo', '.rdr', '.full']:
fbase, fext = os.path.splitext(fbase)
if not fext:
fext = fbase
if metafile0.endswith('.rsc'):
atr.update(read_roipac_rsc(metafile0))
if 'FILE_TYPE' not in atr.keys():
atr['FILE_TYPE'] = fext
elif metafile0.endswith('.xml'):
atr.update(read_isce_xml(metafile0))
if 'FILE_TYPE' not in atr.keys():
atr['FILE_TYPE'] = fext #atr.get('image_type', fext)
elif metafile0.endswith('.par'):
atr.update(read_gamma_par(metafile0))
atr['FILE_TYPE'] = fext
elif metafile0.endswith('.hdr'):
atr.update(read_envi_hdr(metafile0))
atr['FILE_TYPE'] = atr['file type']
# UNIT
k = atr['FILE_TYPE'].replace('.', '')
if k == 'ifgramStack':
if datasetName and datasetName in datasetUnitDict.keys():
atr['UNIT'] = datasetUnitDict[datasetName]
else:
atr['UNIT'] = 'radian'
elif 'UNIT' not in atr.keys():
if datasetName and datasetName in datasetUnitDict.keys():
atr['UNIT'] = datasetUnitDict[datasetName]
elif k in datasetUnitDict.keys():
atr['UNIT'] = datasetUnitDict[k]
else:
atr['UNIT'] = '1'
# FILE_PATH
atr['FILE_PATH'] = os.path.abspath(fname)
if standardize:
atr = standardize_metadata(atr)
return atr
