def _get_ancillary_metadata(mtl_doc,
wo_doc,
mtl_name_offset=None,
order_dir_offset=None,
properties_offsets=None):
#: :type: Path
specified_path = _get_node_text(
order_dir_offset, wo_doc,
Path) if order_dir_offset and wo_doc else None
used_file_name = _get(
mtl_doc, *mtl_name_offset) if mtl_name_offset and mtl_doc else None
# Read any properties of the ancillary file form the MTL.
properties = {}
if mtl_doc and properties_offsets:
for property_name, offset in properties_offsets.items():
val = _get(mtl_doc, *offset)
if val:
properties[property_name] = val
if not specified_path or not specified_path.exists():
_LOG.warning('No path found to locate ancillary file %s',
used_file_name)
# If there's no information of the ancillary, don't bother.
if (not used_file_name) and (not properties):
return None
return ptype.AncillaryMetadata(name=used_file_name,
properties=properties)
if specified_path.is_file():
# They specified an exact file to Pinkmatter rather than a search directory.
file_path = specified_path
else:
# Else is a directory.
# Find the file that was used inside it.
# No file was used by Pinkmatter (Eg. TIRS ancillary in an OLI dataset)
if not used_file_name:
return None
file_path = _get_file(specified_path, used_file_name)
_LOG.info('Found ancillary path %s', file_path)
return ptype.AncillaryMetadata.from_file(file_path, properties=properties)
# ground_control_points_model=47,
# geometric_rmse_model=4.582,
# geometric_rmse_model_x=3.370,
# geometric_rmse_model_y=3.104,
bands={}
),
lineage=ptype.LineageMetadata(
algorithm=ptype.AlgorithmMetadata(
name='LPGS',
version='12.5.0',
parameters={}
),
ancillary_quality='DEFINITIVE',
ancillary={
'cpf': ptype.AncillaryMetadata(
name='L7CPF20050101_20050331.09'
),
# We have the properties (quality) of the ancillary but not the file.
'ephemeris': ptype.AncillaryMetadata(
properties={'type': 'DEFINITIVE'}
)
}
)
)
class TestMtlRead(unittest.TestCase):
def test_ls7_equivalence(self):
assert_expected_mtl(
Path(os.path.join(os.path.dirname(__file__), FILENAME)),
EXPECTED_OUT
def fill_metadata(self, dataset, path, additional_files=()):
"""
:type additional_files: tuple[Path]
:type dataset: ptype.DatasetMetadata
:type path: Path
:rtype: ptype.DatasetMetadata
"""
dataset.ga_level = 'P55'
# Copy relevant fields from source nbar.
if 'nbar' in dataset.lineage.source_datasets:
source_ortho = dataset.lineage.source_datasets['nbar']
borrow_single_sourced_fields(dataset, source_ortho)
# TODO, it'd be better to grab this from the images, but they're generated after
# this code is run. Copying from Source will do for now
dataset.grid_spatial = deepcopy(
dataset.lineage.source_datasets['nbar'].grid_spatial)
contiguous_data_bit = 0b100000000
dataset.grid_spatial.projection.valid_data = self.calculate_valid_data_region(
path, contiguous_data_bit)
dataset.format_ = ptype.FormatMetadata('GeoTIFF')
with open(str(path.joinpath(self.METADATA_FILE))) as f:
pq_metadata = yaml.load(f, Loader=Loader)
if not dataset.lineage:
dataset.lineage = ptype.LineageMetadata()
dataset.lineage.algorithm = ptype.AlgorithmMetadata(
name='pqa',
version=str(
pq_metadata['algorithm_information']['software_version']),
doi=pq_metadata['algorithm_information']['pq_doi'])
# Add ancillary files
ancils = pq_metadata['ancillary']
ancil_files = {}
for name, values in ancils.items():
ancil_files[name] = ptype.AncillaryMetadata(
type_=name,
name=values['data_source'],
uri=values['data_file'],
file_owner=values['user'],
# PyYAML parses these as datetimes already.
access_dt=values['accessed'],
modification_dt=values['modified'])
if ancil_files:
dataset.lineage.ancillary = ancil_files
product_flags = {}
# Record which tests where run in 'product_flags'
for test_name, val in pq_metadata['tests_run'].items():
product_flags['tested_%s' % test_name] = val
dataset.product_flags = product_flags
return dataset
def fill_metadata(self, dataset, path, additional_files=()):
"""
:type additional_files: tuple[Path]
:type dataset: ptype.DatasetMetadata
:type path: Path
:rtype: ptype.DatasetMetadata
"""
with open(str(path.joinpath(self.METADATA_FILE))) as f:
nbar_metadata = yaml.load(f, Loader=Loader)
# Copy relevant fields from source ortho.
if 'level1' in dataset.lineage.source_datasets:
source_ortho = dataset.lineage.source_datasets['level1']
borrow_single_sourced_fields(dataset, source_ortho)
# TODO, it'd be better to grab this from the images, but they're generated after
# this code is run. Copying from Source will do for now
dataset.grid_spatial = deepcopy(
dataset.lineage.source_datasets['level1'].grid_spatial)
dataset.grid_spatial.projection.valid_data = self.calculate_valid_data_region(
path)
if not dataset.lineage:
dataset.lineage = ptype.LineageMetadata()
self._fill_algorithm_information(
dataset, nbar_metadata['algorithm_information'])
dataset.product_doi = nbar_metadata['algorithm_information'][
'arg25_doi']
# Extract ancillary file data and values
parameters = {}
ancils = nbar_metadata['ancillary_data']
brdfs = ancils.pop('brdf', {})
brdf_ancils = {
'_'.join((band_name, 'brdf', ancil_type)): values
for band_name, ancil_types in brdfs.items()
for ancil_type, values in ancil_types.items()
}
ancils.update(brdf_ancils)
# Add algorithm parameters
for name, values in ancils.items():
parameters[name] = values['value']
if parameters:
dataset.lineage.algorithm.parameters = parameters
# Add ancillary files
ancil_files = {}
for name, values in ancils.items():
if 'data_file' not in values:
continue
ancil_files[name] = ptype.AncillaryMetadata(
type_=name,
name=values['data_file'].rpartition('/')[2],
uri=values['data_file'],
file_owner=values['user'],
# PyYAML parses these as datetimes already.
access_dt=values['accessed'],
modification_dt=values['modified'])
if ancil_files:
dataset.lineage.ancillary = ancil_files
# All NBARs are P54. (source: Lan Wei)
dataset.ga_level = 'P54'
dataset.format_ = ptype.FormatMetadata('GeoTIFF')
return dataset
def _build_ls8_ortho():
_reset_runtime_id()
return ptype.DatasetMetadata(
id_=uuid.UUID('17b92c16-51d3-11e4-909d-005056bb6972'),
ga_label='LS8_OLITIRS_OTH_P51_GALPGS01-002_101_078_20141012',
product_type='GAORTHO01',
creation_dt=dateutil.parser.parse('2014-10-12 05:46:20'),
size_bytes=2386550 * 1024,
platform=ptype.PlatformMetadata(code='LANDSAT-8'),
instrument=ptype.InstrumentMetadata(name='OLI_TIRS',
type_="Multi-Spectral",
operation_mode='PUSH-BROOM'),
format_=ptype.FormatMetadata(name='GeoTiff', version=1),
extent=ptype.ExtentMetadata(
reference_system='WGS84',
coord=ptype.CoordPolygon(ul=ptype.Coord(lat=-24.97, lon=133.97969),
ur=ptype.Coord(lat=-24.96826,
lon=136.24838),
lr=ptype.Coord(lat=-26.96338,
lon=136.26962),
ll=ptype.Coord(lat=-26.96528,
lon=133.96233)),
from_dt=dateutil.parser.parse("2014-10-12T00:55:54"),
center_dt=dateutil.parser.parse("2014-10-12T00:56:06"),
to_dt=dateutil.parser.parse("2014-10-12T00:56:18"),
),
grid_spatial=ptype.GridSpatialMetadata(
dimensions=[
ptype.DimensionMetadata(name='sample',
resolution=25.0,
size=9161),
ptype.DimensionMetadata(name='line',
resolution=25.0,
size=9161)
],
projection=ptype.ProjectionMetadata(
centre_point=ptype.Point(511512.500000, 7127487.500000),
geo_ref_points=ptype.PointPolygon(
ul=ptype.Point(397012.5, 7237987.5),
ur=ptype.Point(626012.5, 7237987.5),
ll=ptype.Point(397012.5, 7016987.5),
lr=ptype.Point(626012.5, 7016987.5)),
datum='GDA94',
ellipsoid='GRS80',
point_in_pixel='UL',
map_projection='UTM',
resampling_option='CUBIC_CONVOLUTION',
zone=-53)),
browse={
'medium':
ptype.BrowseMetadata(path=Path(
'product/LS8_OLITIRS_OTH_P51_GALPGS01-032_101_078_20141012.jpg'
),
file_type='image/jpg',
cell_size=219.75,
red_band=7,
green_band=5,
blue_band=1),
'full':
ptype.BrowseMetadata(path=Path(
'LS8_OLITIRS_OTH_P51_GALPGS01-032_101_078_20141012_FR.jpg'),
file_type='image/jpg',
cell_size=25.0,
red_band=7,
green_band=5,
blue_band=1)
},
image=ptype.ImageMetadata(
satellite_ref_point_start=ptype.Point(101, 78),
cloud_cover_percentage=0,
cloud_cover_details=None,
sun_elevation=58.00268508,
sun_azimuth=59.41814014,
ground_control_points_model=420,
geometric_rmse_model=4.610,
geometric_rmse_model_x=3.527,
geometric_rmse_model_y=2.968,
# TODO: What are these two?
viewing_incidence_angle_long_track=0,
viewing_incidence_angle_x_track=0,
bands={
'coastal_aerosol':
ptype.BandMetadata(
path=Path('product/LC81010782014285LGN00_B1.TIF'),
number=1,
type_='reflective',
cell_size=25.0,
),
'visible_blue':
ptype.BandMetadata(
path=Path('product/LC81010782014285LGN00_B2.TIF'),
number=2,
type_='reflective',
cell_size=25.0,
),
'visible_green':
ptype.BandMetadata(
path=Path('product/LC81010782014285LGN00_B3.TIF'),
number=3,
type_='reflective',
cell_size=25.0,
),
'visible_red':
ptype.BandMetadata(
path=Path('product/LC81010782014285LGN00_B4.TIF'),
number=4,
type_='reflective',
cell_size=25.0,
),
'near_infrared':
ptype.BandMetadata(
path=Path('product/LC81010782014285LGN00_B5.TIF'),
number=5,
type_='reflective',
cell_size=25.0,
),
'short_wave_infrared1':
ptype.BandMetadata(
path=Path('product/LC81010782014285LGN00_B6.TIF'),
number=6,
type_='reflective',
cell_size=25.0,
),
'short_wave_infrared2':
ptype.BandMetadata(
path=Path('product/LC81010782014285LGN00_B7.TIF'),
number=7,
type_='reflective',
cell_size=25.0,
),
'panchromatic':
ptype.BandMetadata(
path=Path('product/LC81010782014285LGN00_B8.TIF'),
number=8,
type_='panchromatic',
cell_size=12.50,
shape=ptype.Point(17761, 18241),
),
'cirrus':
ptype.BandMetadata(
path=Path('product/LC81010782014285LGN00_B9.TIF'),
number=9,
type_='atmosphere',
),
'thermal_infrared1':
ptype.BandMetadata(
path=Path('product/LC81010782014285LGN00_B10.TIF'),
number=10,
type_='thermal',
cell_size=25.0,
shape=ptype.Point(8881, 9121),
),
'thermal_infrared2':
ptype.BandMetadata(
path=Path('product/LC81010782014285LGN00_B11.TIF'),
number=11,
type_='thermal',
cell_size=25.0,
shape=ptype.Point(8881, 9121),
),
'quality':
ptype.BandMetadata(
path=Path('product/LC81010782014285LGN00_BQA.TIF'),
number='QA',
type_='quality',
)
}),
lineage=ptype.LineageMetadata(
algorithm=ptype.AlgorithmMetadata(
name='Pinkmatter Landsat Processor',
version='3.3.3104',
parameters={
'resampling': 'CC',
'radiometric_correction': 'CPF',
'orientation': 'NUP',
'hemisphere': 'S',
}),
machine=ptype.MachineMetadata(
hostname='rhe-jm-prod08.prod.lan',
type_id='jobmanager',
uname=
'Linux rhe-jm-dev08.dev.lan 2.6.32-279.22.1.el6.x86_64 #1 SMP Sun Oct '
'12 '
'09:21:40 EST 2014 x86_64 x86_64 x86_64 GNU/Linux'),
ancillary={
'cpf':
ptype.AncillaryMetadata(
name='L8CPF20141001_20141231.01',
uri=
'/eoancillarydata/sensor-specific/LANDSAT8/CalibrationParameterFile'
'/L8CPF20141001_20141231.01'),
'bpf_tirs':
ptype.AncillaryMetadata(
name='LT8BPF20141012002432_20141012020301.01',
uri=
'/eoancillarydata/sensor-specific/LANDSAT8/BiasParameterFile/2014/10'
'/LT8BPF20141012002432_20141012020301.01'),
'bpf_oli':
ptype.AncillaryMetadata(
name='LO8BPF20141012002825_20141012011100.01',
uri=
'/eoancillarydata/sensor-specific/LANDSAT8/BiasParameterFile/2014/10'
'/LT8BPF20141012002432_20141012020301.01'),
'rlut':
ptype.AncillaryMetadata(name='L8RLUT20130211_20431231v09.h5')
},
source_datasets={'satellite_telemetry_data': _build_ls8_raw()}))
cloud_cover_percentage=0.01,
sun_azimuth=59.57807899,
sun_elevation=57.89670734,
sun_earth_distance=0.998137,
ground_control_points_model=420,
geometric_rmse_model=4.61,
geometric_rmse_model_x=2.968,
geometric_rmse_model_y=3.527,
bands={}),
lineage=ptype.LineageMetadata(
algorithm=ptype.AlgorithmMetadata(name='LPGS',
version='2.3.0',
parameters={}),
ancillary={
'rlut':
ptype.AncillaryMetadata(name='L8RLUT20130211_20431231v09.h5'),
'bpf_tirs':
ptype.AncillaryMetadata(
name='LT8BPF20141012002432_20141012011154.02'),
'bpf_oli':
ptype.AncillaryMetadata(
name='LO8BPF20141012002825_20141012011100.01'),
'cpf':
ptype.AncillaryMetadata(name='L8CPF20141001_20141231.01')
}))
class TestMtlRead(unittest.TestCase):
def test_ls8_equivalence(self):
assert_expected_mtl(
Path(os.path.join(os.path.dirname(__file__), FILENAME)),