class SentinelDownloader(object):
def __init__(self, user, password, api_url='https://scihub.copernicus.eu/apihub'):
self._apiname = api_url
self._user = user
self._password = password
# init logger
root = logging.getLogger()
root.addHandler(logging.StreamHandler(
sys.stderr
))
if self._apiname == 'https://scihub.copernicus.eu/apihub':
try:
from sentinelsat import SentinelAPI
except ImportError as e:
gs.fatal(_("Module requires sentinelsat library: {}").format(e))
# connect SciHub via API
self._api = SentinelAPI(self._user, self._password,
api_url=self._apiname
)
elif self._apiname == 'USGS_EE':
try:
import landsatxplore.api
from landsatxplore.errors import EarthExplorerError
except ImportError as e:
gs.fatal(_("Module requires landsatxplore library: {}").format(e))
api_login = False
while api_login is False:
# avoid login conflict in possible parallel execution
try:
self._api = landsatxplore.api.API(self._user,
self._password)
api_login = True
except EarthExplorerError as e:
time.sleep(1)
self._products_df_sorted = None
def filter(self, area, area_relation,
clouds=None, producttype=None, limit=None, query={},
start=None, end=None, sortby=[], asc=True, relativeorbitnumber=None):
args = {}
if clouds:
args['cloudcoverpercentage'] = (0, int(clouds))
if relativeorbitnumber:
args['relativeorbitnumber'] = relativeorbitnumber
if producttype.startswith('S2') and int(relativeorbitnumber) > 143:
gs.warning("This relative orbit number is out of range")
elif int(relativeorbitnumber) > 175:
gs.warning(_("This relative orbit number is out of range"))
if producttype:
args['producttype'] = producttype
if producttype.startswith('S2'):
args['platformname'] = 'Sentinel-2'
else:
args['platformname'] = 'Sentinel-1'
if not start:
start = 'NOW-60DAYS'
else:
start = start.replace('-', '')
if not end:
end = 'NOW'
else:
end = end.replace('-', '')
if query:
redefined = [value for value in args.keys() if value in query.keys()]
if redefined:
gs.warning(_("Query overrides already defined options ({})").format(
','.join(redefined)
))
args.update(query)
gs.verbose(_("Query: area={} area_relation={} date=({}, {}) args={}").format(
area, area_relation, start, end, args
))
products = self._api.query(
area=area, area_relation=area_relation,
date=(start, end),
**args
)
products_df = self._api.to_dataframe(products)
if len(products_df) < 1:
gs.message(_("No product found"))
return
# sort and limit to first sorted product
if sortby:
self._products_df_sorted = products_df.sort_values(
sortby,
ascending=[asc] * len(sortby)
)
else:
self._products_df_sorted = products_df
if limit:
self._products_df_sorted = self._products_df_sorted.head(int(limit))
gs.message(_("{} Sentinel product(s) found").format(len(self._products_df_sorted)))
def list(self):
if self._products_df_sorted is None:
return
id_kw = ('uuid', 'entity_id')
identifier_kw = ('identifier', 'display_id')
cloud_kw = ('cloudcoverpercentage', 'cloud_cover')
time_kw = ('beginposition', 'acquisition_date')
kw_idx = 1 if self._apiname == 'USGS_EE' else 0
for idx in range(len(self._products_df_sorted[id_kw[kw_idx]])):
if cloud_kw[kw_idx] in self._products_df_sorted:
ccp = '{0:2.0f}%'.format(
float(self._products_df_sorted[cloud_kw[kw_idx]][idx]))
else:
ccp = 'cloudcover_NA'
print_str = '{0} {1}'.format(
self._products_df_sorted[id_kw[kw_idx]][idx],
self._products_df_sorted[identifier_kw[kw_idx]][idx])
if kw_idx == 1:
time_string = self._products_df_sorted[time_kw[kw_idx]][idx]
else:
time_string = self._products_df_sorted[
time_kw[kw_idx]][idx].strftime("%Y-%m-%dT%H:%M:%SZ")
print_str += ' {0} {1}'.format(time_string, ccp)
if kw_idx == 0:
print_str += ' {0}'.format(
self._products_df_sorted['producttype'][idx])
print(print_str)
def download(self, output, sleep=False, maxretry=False,
datasource='ESA_COAH'):
if self._products_df_sorted is None:
return
create_dir(output)
gs.message(_("Downloading data into <{}>...").format(output))
if datasource == 'USGS_EE':
from landsatxplore.earthexplorer import EarthExplorer
from landsatxplore.errors import EarthExplorerError
from zipfile import ZipFile
ee_login = False
while ee_login is False:
# avoid login conflict in possible parallel execution
try:
ee = EarthExplorer(self._user, self._password)
ee_login = True
except EarthExplorerError as e:
time.sleep(1)
for idx in range(len(self._products_df_sorted['entity_id'])):
scene = self._products_df_sorted['entity_id'][idx]
identifier = self._products_df_sorted['display_id'][idx]
zip_file = os.path.join(output, '{}.zip'.format(identifier))
gs.message(_("Downloading {}...").format(identifier))
try:
ee.download(identifier=identifier, output_dir=output, timeout=600)
except EarthExplorerError as e:
gs.fatal(_(e))
ee.logout()
# extract .zip to get "usual" .SAFE
with ZipFile(zip_file, 'r') as zip:
safe_name = zip.namelist()[0].split('/')[0]
outpath = os.path.join(output, safe_name)
zip.extractall(path=output)
gs.message(_("Downloaded to <{}>").format(outpath))
try:
os.remove(zip_file)
except Exception as e:
gs.warning(_("Unable to remove {0}:{1}").format(
zip_file, e))
elif datasource == "ESA_COAH":
for idx in range(len(self._products_df_sorted['uuid'])):
gs.message('{} -> {}.SAFE'.format(
self._products_df_sorted['uuid'][idx],
os.path.join(output, self._products_df_sorted['identifier'][idx])
))
# download
out = self._api.download(self._products_df_sorted['uuid'][idx],
output)
if sleep:
x = 1
online = out['Online']
while not online:
# sleep is in minutes so multiply by 60
time.sleep(int(sleep) * 60)
out = self._api.download(self._products_df_sorted['uuid'][idx],
output)
x += 1
if x > maxretry:
online = True
elif datasource == 'GCS':
for scene_id in self._products_df_sorted['identifier']:
gs.message(_("Downloading {}...").format(scene_id))
dl_code = download_gcs(scene_id, output)
if dl_code == 0:
gs.message(_("Downloaded to {}").format(
os.path.join(output, '{}.SAFE'.format(scene_id))))
else:
# remove incomplete file
del_folder = os.path.join(output,
'{}.SAFE'.format(scene_id))
try:
shutil.rmtree(del_folder)
except Exception as e:
gs.warning(_("Unable to removed unfinished "
"download {}".format(del_folder)))
def save_footprints(self, map_name):
if self._products_df_sorted is None:
return
if self._apiname == 'USGS_EE':
gs.fatal(_(
"USGS Earth Explorer does not support footprint download."))
try:
from osgeo import ogr, osr
except ImportError as e:
gs.fatal(_("Option <footprints> requires GDAL library: {}").format(e))
gs.message(_("Writing footprints into <{}>...").format(map_name))
driver = ogr.GetDriverByName("GPKG")
tmp_name = gs.tempfile() + '.gpkg'
data_source = driver.CreateDataSource(tmp_name)
srs = osr.SpatialReference()
srs.ImportFromEPSG(4326)
# features can be polygons or multi-polygons
layer = data_source.CreateLayer(str(map_name), srs, ogr.wkbMultiPolygon)
# attributes
attrs = OrderedDict([
("uuid", ogr.OFTString),
("ingestiondate", ogr.OFTString),
("cloudcoverpercentage", ogr.OFTInteger),
("producttype", ogr.OFTString),
("identifier", ogr.OFTString)
])
# Sentinel-1 data does not have cloudcoverpercentage
prod_types = [type for type in self._products_df_sorted["producttype"]]
s1_types = ["SLC", "GRD"]
if any(type in prod_types for type in s1_types):
del attrs["cloudcoverpercentage"]
for key in attrs.keys():
field = ogr.FieldDefn(key, attrs[key])
layer.CreateField(field)
# features
for idx in range(len(self._products_df_sorted['uuid'])):
wkt = self._products_df_sorted['footprint'][idx]
feature = ogr.Feature(layer.GetLayerDefn())
newgeom = ogr.CreateGeometryFromWkt(wkt)
# convert polygons to multi-polygons
newgeomtype = ogr.GT_Flatten(newgeom.GetGeometryType())
if newgeomtype == ogr.wkbPolygon:
multigeom = ogr.Geometry(ogr.wkbMultiPolygon)
multigeom.AddGeometryDirectly(newgeom)
feature.SetGeometry(multigeom)
else:
feature.SetGeometry(newgeom)
for key in attrs.keys():
if key == 'ingestiondate':
value = self._products_df_sorted[key][idx].strftime("%Y-%m-%dT%H:%M:%SZ")
else:
value = self._products_df_sorted[key][idx]
feature.SetField(key, value)
layer.CreateFeature(feature)
feature = None
data_source = None
# coordinates of footprints are in WKT -> fp precision issues
# -> snap
gs.run_command('v.import', input=tmp_name, output=map_name,
layer=map_name, snap=1e-10, quiet=True
)
def get_products_from_uuid_usgs(self, uuid_list):
scenes = []
for uuid in uuid_list:
metadata = self._api.metadata(uuid, 'SENTINEL_2A')
scenes.append(metadata)
scenes_df = pandas.DataFrame.from_dict(scenes)
self._products_df_sorted = scenes_df
gs.message(_("{} Sentinel product(s) found").format(
len(self._products_df_sorted)))
def set_uuid(self, uuid_list):
"""Set products by uuid.
TODO: Find better implementation
:param uuid: uuid to download
"""
if self._apiname == 'USGS_EE':
self.get_products_from_uuid_usgs(uuid_list)
else:
from sentinelsat.sentinel import SentinelAPIError
self._products_df_sorted = {'uuid': []}
for uuid in uuid_list:
try:
odata = self._api.get_product_odata(uuid, full=True)
except SentinelAPIError as e:
gs.error(_("{0}. UUID {1} skipped".format(e, uuid)))
continue
for k, v in odata.items():
if k == 'id':
k = 'uuid'
elif k == 'Sensing start':
k = 'beginposition'
elif k == 'Product type':
k = 'producttype'
elif k == 'Cloud cover percentage':
k = 'cloudcoverpercentage'
elif k == 'Identifier':
k = 'identifier'
elif k == 'Ingestion Date':
k = 'ingestiondate'
elif k == 'footprint':
pass
else:
continue
if k not in self._products_df_sorted:
self._products_df_sorted[k] = []
self._products_df_sorted[k].append(v)
def filter_USGS(self, area, area_relation, clouds=None, producttype=None,
limit=None, query={}, start=None, end=None, sortby=[],
asc=True, relativeorbitnumber=None):
if area_relation != 'Intersects':
gs.fatal(_(
"USGS Earth Explorer only supports area_relation"
" 'Intersects'"))
if relativeorbitnumber:
gs.fatal(_(
"USGS Earth Explorer does not support 'relativeorbitnumber'"
" option."))
if producttype and producttype != 'S2MSI1C':
gs.fatal(_(
"USGS Earth Explorer only supports producttype S2MSI1C"))
if query:
if not any(key in query for key in ['identifier', 'filename',
'usgs_identifier']):
gs.fatal(_(
"USGS Earth Explorer only supports query options"
" 'filename', 'identifier' or 'usgs_identifier'."))
if 'usgs_identifier' in query:
# get entityId from usgs identifier and directly save results
usgs_id = query['usgs_identifier']
check_s2l1c_identifier(usgs_id, source='usgs')
# entity_id = self._api.lookup('SENTINEL_2A', [usgs_id],
# inverse=True)
entity_id = self._api.get_entity_id([usgs_id], 'SENTINEL_2A')
self.get_products_from_uuid_usgs(entity_id)
return
else:
if "filename" in query:
esa_id = query['filename'].replace('.SAFE', '')
else:
esa_id = query['identifier']
check_s2l1c_identifier(esa_id, source='esa')
esa_prod_id = esa_id.split('_')[-1]
utm_tile = esa_id.split('_')[-2]
acq_date = esa_id.split('_')[2].split('T')[0]
acq_date_string = '{0}-{1}-{2}'.format(
acq_date[:4], acq_date[4:6], acq_date[6:])
start_date = end_date = acq_date_string
# build the USGS style S2-identifier
if utm_tile.startswith('T'):
utm_tile_base = utm_tile[1:]
bbox = get_bbox_from_S2_UTMtile(utm_tile_base)
else:
# get coordinate pairs from wkt string
str_1 = 'POLYGON(('
str_2 = '))'
coords = area[area.find(str_1)+len(str_1):area.rfind(str_2)].split(',')
# add one space to first pair for consistency
coords[0] = ' ' + coords[0]
lons = [float(pair.split(' ')[1]) for pair in coords]
lats = [float(pair.split(' ')[2]) for pair in coords]
bbox = (min(lons), min(lats), max(lons), max(lats))
start_date = start
end_date = end
usgs_args = {
'dataset': 'SENTINEL_2A',
'bbox': bbox,
'start_date': start_date,
'end_date': end_date
}
if clouds:
usgs_args['max_cloud_cover'] = clouds
if limit:
usgs_args['max_results'] = limit
scenes = self._api.search(**usgs_args)
self._api.logout()
if query:
# check if the UTM-Tile is correct, remove otherwise
for scene in scenes:
if scene['display_id'].split('_')[1] != utm_tile:
scenes.remove(scene)
# remove redundant scene
if len(scenes) == 2:
for scene in scenes:
prod_id = scene['display_id'].split('_')[-1]
if prod_id != esa_prod_id:
scenes.remove(scene)
if len(scenes) < 1:
gs.message(_("No product found"))
return
scenes_df = pandas.DataFrame.from_dict(scenes)
if sortby:
# replace sortby keywords with USGS keywords
for idx, keyword in enumerate(sortby):
if keyword == 'cloudcoverpercentage':
sortby[idx] = 'cloud_cover'
# turn cloudcover to float to make it sortable
scenes_df['cloud_cover'] = pandas.to_numeric(
scenes_df['cloud_cover'])
elif keyword == 'ingestiondate':
sortby[idx] = 'acquisition_date'
# what does sorting by footprint mean
elif keyword == 'footprint':
sortby[idx] = 'display_id'
self._products_df_sorted = scenes_df.sort_values(
sortby,
ascending=[asc] * len(sortby), ignore_index=True
)
else:
self._products_df_sorted = scenes_df
gs.message(_("{} Sentinel product(s) found").format(
len(self._products_df_sorted)))
class SentinelDownloader(object):
def __init__(
self,
user,
password,
api_url="https://apihub.copernicus.eu/apihub",
cred_req=True,
):
self._apiname = api_url
self._user = user
self._password = password
self._cred_req = cred_req
# init logger
root = logging.getLogger()
root.addHandler(logging.StreamHandler(sys.stderr))
if self._apiname not in ["USGS_EE", "GCS"]:
try:
from sentinelsat import SentinelAPI
except ImportError as e:
gs.fatal(
_("Module requires sentinelsat library: {}").format(e))
# connect SciHub via API
self._api = SentinelAPI(self._user,
self._password,
api_url=self._apiname)
elif self._apiname == "USGS_EE":
api_login = False
while api_login is False:
# avoid login conflict in possible parallel execution
try:
self._api = landsatxplore.api.API(self._user,
self._password)
api_login = True
except EarthExplorerError as e:
time.sleep(1)
self._products_df_sorted = None
def filter(
self,
area,
area_relation,
clouds=None,
producttype=None,
limit=None,
query={},
start=None,
end=None,
sortby=[],
asc=True,
relativeorbitnumber=None,
):
# Dict to identify plaforms from requested product
platforms = {
"SL": "Sentinel-1",
"GR": "Sentinel-1",
"OC": "Sentinel-1",
"S2": "Sentinel-2",
"S3": "Sentinel-3",
}
args = {}
if clouds:
args["cloudcoverpercentage"] = (0, int(clouds))
if relativeorbitnumber:
args["relativeorbitnumber"] = relativeorbitnumber
if producttype.startswith("S2") and int(relativeorbitnumber) > 143:
gs.warning("This relative orbit number is out of range")
elif int(relativeorbitnumber) > 175:
gs.warning(_("This relative orbit number is out of range"))
if producttype:
if producttype.startswith("S3"):
# Using custom product names for Sentinel-3 products that look less cryptic
split = [0, 2, 4, 5, 8]
args["producttype"] = "_".join([
producttype[i:j] for i, j in zip(split, split[1:] + [None])
][1:]).ljust(11, "_")
else:
args["producttype"] = producttype
args["platformname"] = platforms[producttype[0:2]]
if not start:
start = "NOW-60DAYS"
else:
start = start.replace("-", "")
if not end:
end = "NOW"
else:
end = end.replace("-", "")
if query:
redefined = [
value for value in args.keys() if value in query.keys()
]
if redefined:
gs.warning(
_("Query overrides already defined options ({})").format(
",".join(redefined)))
args.update(query)
gs.verbose(
_("Query: area={} area_relation={} date=({}, {}) args={}").format(
area, area_relation, start, end, args))
if self._cred_req is False:
# in the main function it is ensured that there is an "identifier" query
self._products_df_sorted = pandas.DataFrame(
{"identifier": [query["identifier"]]})
return
products = self._api.query(area=area,
area_relation=area_relation,
date=(start, end),
**args)
products_df = self._api.to_dataframe(products)
if len(products_df) < 1:
gs.message(_("No product found"))
return
# sort and limit to first sorted product
if sortby:
self._products_df_sorted = products_df.sort_values(
sortby, ascending=[asc] * len(sortby))
else:
self._products_df_sorted = products_df
if limit:
self._products_df_sorted = self._products_df_sorted.head(
int(limit))
gs.message(
_("{} Sentinel product(s) found").format(
len(self._products_df_sorted)))
def list(self):
if self._products_df_sorted is None:
return
id_kw = ("uuid", "entity_id")
identifier_kw = ("identifier", "display_id")
cloud_kw = ("cloudcoverpercentage", "cloud_cover")
time_kw = ("beginposition", "acquisition_date")
kw_idx = 1 if self._apiname == "USGS_EE" else 0
for idx in range(len(self._products_df_sorted[id_kw[kw_idx]])):
if cloud_kw[kw_idx] in self._products_df_sorted:
ccp = "{0:2.0f}%".format(
float(self._products_df_sorted[cloud_kw[kw_idx]][idx]))
else:
ccp = "cloudcover_NA"
print_str = "{0} {1}".format(
self._products_df_sorted[id_kw[kw_idx]][idx],
self._products_df_sorted[identifier_kw[kw_idx]][idx],
)
if kw_idx == 1:
time_string = self._products_df_sorted[time_kw[kw_idx]][idx]
else:
time_string = self._products_df_sorted[
time_kw[kw_idx]][idx].strftime("%Y-%m-%dT%H:%M:%SZ")
print_str += " {0} {1}".format(time_string, ccp)
if kw_idx == 0:
print_str += " {0}".format(
self._products_df_sorted["producttype"][idx])
print_str += " {0}".format(
self._products_df_sorted["size"][idx])
print(print_str)
def skip_existing(self, output, pattern_file):
prod_df_type = type(self._products_df_sorted)
# Check i skipping is possible/required
if prod_df_type != dict:
if self._products_df_sorted.empty:
return
elif not self._products_df_sorted or os.path.exists(output) == False:
return
# Check if ingestion date is returned by API
if "ingestiondate" not in self._products_df_sorted:
gs.warning(
_("Ingestiondate not returned. Cannot filter previously downloaded scenes"
))
return
# Check for previously downloaded scenes
existing_files = [
f for f in os.listdir(output)
if re.search(r".zip$|.safe$|.ZIP$|.SAFE$", f)
]
if len(existing_files) <= 1:
return
# Filter by ingestion date
skiprows = []
for idx, display_id in enumerate(
self._products_df_sorted["identifier"]):
existing_file = [
sfile for sfile in existing_files if display_id in sfile
]
if existing_file:
creation_time = datetime.fromtimestamp(
os.path.getctime(existing_file[0]))
if self._products_df_sorted["ingestiondate"][
idx] <= creation_time:
gs.message(
_("Skipping scene: {} which is already downloaded.".
format(self._products_df_sorted["identifier"][idx])))
skiprows.append(display_id)
if prod_df_type == dict:
for scene in skiprows:
idx = self._products_df_sorted["identifier"].index(scene)
for key in self._products_df_sorted:
self._products_df_sorted[key].pop(idx)
else:
self._products_df_sorted = self._products_df_sorted[
~self._products_df_sorted["identifier"].isin(skiprows)]
def download(self,
output,
sleep=False,
maxretry=False,
datasource="ESA_COAH"):
if self._products_df_sorted is None:
return
create_dir(output)
gs.message(_("Downloading data into <{}>...").format(output))
if datasource == "USGS_EE":
from landsatxplore.earthexplorer import EarthExplorer
from landsatxplore.errors import EarthExplorerError
from zipfile import ZipFile
ee_login = False
while ee_login is False:
# avoid login conflict in possible parallel execution
try:
ee = EarthExplorer(self._user, self._password)
ee_login = True
except EarthExplorerError as e:
time.sleep(1)
for idx in range(len(self._products_df_sorted["entity_id"])):
scene = self._products_df_sorted["entity_id"][idx]
identifier = self._products_df_sorted["display_id"][idx]
zip_file = os.path.join(output, "{}.zip".format(identifier))
gs.message(_("Downloading {}...").format(identifier))
try:
ee.download(identifier=identifier,
output_dir=output,
timeout=600)
except EarthExplorerError as e:
gs.fatal(_(e))
ee.logout()
# extract .zip to get "usual" .SAFE
with ZipFile(zip_file, "r") as zip:
safe_name = zip.namelist()[0].split("/")[0]
outpath = os.path.join(output, safe_name)
zip.extractall(path=output)
gs.message(_("Downloaded to <{}>").format(outpath))
try:
os.remove(zip_file)
except Exception as e:
gs.warning(
_("Unable to remove {0}:{1}").format(zip_file, e))
elif datasource == "ESA_COAH":
for idx in range(len(self._products_df_sorted["uuid"])):
gs.message("{} -> {}.SAFE".format(
self._products_df_sorted["uuid"][idx],
os.path.join(output,
self._products_df_sorted["identifier"][idx]),
))
# download
out = self._api.download(self._products_df_sorted["uuid"][idx],
output)
if sleep:
x = 1
online = out["Online"]
while not online:
# sleep is in minutes so multiply by 60
time.sleep(int(sleep) * 60)
out = self._api.download(
self._products_df_sorted["uuid"][idx], output)
x += 1
if x > maxretry:
online = True
elif datasource == "GCS":
for scene_id in self._products_df_sorted["identifier"]:
gs.message(_("Downloading {}...").format(scene_id))
dl_code = download_gcs(scene_id, output)
if dl_code == 0:
gs.message(
_("Downloaded to {}").format(
os.path.join(output, "{}.SAFE".format(scene_id))))
else:
# remove incomplete file
del_folder = os.path.join(output,
"{}.SAFE".format(scene_id))
try:
shutil.rmtree(del_folder)
except Exception as e:
gs.warning(
_("Unable to removed unfinished "
"download {}".format(del_folder)))
def save_footprints(self, map_name):
if self._products_df_sorted is None:
return
if self._apiname == "USGS_EE":
gs.fatal(
_("USGS Earth Explorer does not support footprint download."))
try:
from osgeo import ogr, osr
except ImportError as e:
gs.fatal(
_("Option <footprints> requires GDAL library: {}").format(e))
gs.message(_("Writing footprints into <{}>...").format(map_name))
driver = ogr.GetDriverByName("GPKG")
tmp_name = gs.tempfile() + ".gpkg"
data_source = driver.CreateDataSource(tmp_name)
srs = osr.SpatialReference()
srs.ImportFromEPSG(4326)
# features can be polygons or multi-polygons
layer = data_source.CreateLayer(str(map_name), srs,
ogr.wkbMultiPolygon)
# attributes
attrs = OrderedDict([
("uuid", ogr.OFTString),
("ingestiondate", ogr.OFTString),
("cloudcoverpercentage", ogr.OFTInteger),
("producttype", ogr.OFTString),
("identifier", ogr.OFTString),
])
# Sentinel-1 data does not have cloudcoverpercentage
prod_types = [type for type in self._products_df_sorted["producttype"]]
if not any(type in prod_types for type in cloudcover_products):
del attrs["cloudcoverpercentage"]
for key in attrs.keys():
field = ogr.FieldDefn(key, attrs[key])
layer.CreateField(field)
# features
for idx in range(len(self._products_df_sorted["uuid"])):
wkt = self._products_df_sorted["footprint"][idx]
feature = ogr.Feature(layer.GetLayerDefn())
newgeom = ogr.CreateGeometryFromWkt(wkt)
# convert polygons to multi-polygons
newgeomtype = ogr.GT_Flatten(newgeom.GetGeometryType())
if newgeomtype == ogr.wkbPolygon:
multigeom = ogr.Geometry(ogr.wkbMultiPolygon)
multigeom.AddGeometryDirectly(newgeom)
feature.SetGeometry(multigeom)
else:
feature.SetGeometry(newgeom)
for key in attrs.keys():
if key == "ingestiondate":
value = self._products_df_sorted[key][idx].strftime(
"%Y-%m-%dT%H:%M:%SZ")
else:
value = self._products_df_sorted[key][idx]
feature.SetField(key, value)
layer.CreateFeature(feature)
feature = None
data_source = None
# coordinates of footprints are in WKT -> fp precision issues
# -> snap
gs.run_command(
"v.import",
input=tmp_name,
output=map_name,
layer=map_name,
snap=1e-10,
quiet=True,
)
def get_products_from_uuid_usgs(self, uuid_list):
scenes = []
for uuid in uuid_list:
metadata = self._api.metadata(uuid, "SENTINEL_2A")
scenes.append(metadata)
scenes_df = pandas.DataFrame.from_dict(scenes)
self._products_df_sorted = scenes_df
gs.message(
_("{} Sentinel product(s) found").format(
len(self._products_df_sorted)))
def set_uuid(self, uuid_list):
"""Set products by uuid.
TODO: Find better implementation
:param uuid: uuid to download
"""
if self._apiname == "USGS_EE":
self.get_products_from_uuid_usgs(uuid_list)
else:
from sentinelsat.sentinel import SentinelAPIError
self._products_df_sorted = {"uuid": []}
for uuid in uuid_list:
try:
odata = self._api.get_product_odata(uuid, full=True)
except SentinelAPIError as e:
gs.error(_("{0}. UUID {1} skipped".format(e, uuid)))
continue
for k, v in odata.items():
if k == "id":
k = "uuid"
elif k == "Sensing start":
k = "beginposition"
elif k == "Product type":
k = "producttype"
elif k == "Cloud cover percentage":
k = "cloudcoverpercentage"
elif k == "Identifier":
k = "identifier"
elif k == "Ingestion Date":
k = "ingestiondate"
elif k == "footprint":
pass
else:
continue
if k not in self._products_df_sorted:
self._products_df_sorted[k] = []
self._products_df_sorted[k].append(v)
def filter_USGS(
self,
area,
area_relation,
clouds=None,
producttype=None,
limit=None,
query={},
start=None,
end=None,
sortby=[],
asc=True,
relativeorbitnumber=None,
):
if area_relation != "Intersects":
gs.fatal(
_("USGS Earth Explorer only supports area_relation"
" 'Intersects'"))
if relativeorbitnumber:
gs.fatal(
_("USGS Earth Explorer does not support 'relativeorbitnumber'"
" option."))
if producttype and producttype != "S2MSI1C":
gs.fatal(
_("USGS Earth Explorer only supports producttype S2MSI1C"))
if query:
if not any(
key in query
for key in ["identifier", "filename", "usgs_identifier"]):
gs.fatal(
_("USGS Earth Explorer only supports query options"
" 'filename', 'identifier' or 'usgs_identifier'."))
if "usgs_identifier" in query:
# get entityId from usgs identifier and directly save results
usgs_id = query["usgs_identifier"]
check_s2l1c_identifier(usgs_id, source="usgs")
# entity_id = self._api.lookup('SENTINEL_2A', [usgs_id],
# inverse=True)
entity_id = self._api.get_entity_id([usgs_id], "SENTINEL_2A")
self.get_products_from_uuid_usgs(entity_id)
return
else:
if "filename" in query:
esa_id = query["filename"].replace(".SAFE", "")
else:
esa_id = query["identifier"]
check_s2l1c_identifier(esa_id, source="esa")
esa_prod_id = esa_id.split("_")[-1]
utm_tile = esa_id.split("_")[-2]
acq_date = esa_id.split("_")[2].split("T")[0]
acq_date_string = "{0}-{1}-{2}".format(acq_date[:4],
acq_date[4:6],
acq_date[6:])
start_date = end_date = acq_date_string
# build the USGS style S2-identifier
if utm_tile.startswith("T"):
utm_tile_base = utm_tile[1:]
bbox = get_bbox_from_S2_UTMtile(utm_tile_base)
else:
# get coordinate pairs from wkt string
str_1 = "POLYGON(("
str_2 = "))"
coords = area[area.find(str_1) +
len(str_1):area.rfind(str_2)].split(",")
# add one space to first pair for consistency
coords[0] = " " + coords[0]
lons = [float(pair.split(" ")[1]) for pair in coords]
lats = [float(pair.split(" ")[2]) for pair in coords]
bbox = (min(lons), min(lats), max(lons), max(lats))
start_date = start
end_date = end
usgs_args = {
"dataset": "SENTINEL_2A",
"bbox": bbox,
"start_date": start_date,
"end_date": end_date,
}
if clouds:
usgs_args["max_cloud_cover"] = clouds
if limit:
usgs_args["max_results"] = limit
scenes = self._api.search(**usgs_args)
self._api.logout()
if query:
# check if the UTM-Tile is correct, remove otherwise
for scene in scenes:
if scene["display_id"].split("_")[1] != utm_tile:
scenes.remove(scene)
# remove redundant scene
if len(scenes) == 2:
for scene in scenes:
prod_id = scene["display_id"].split("_")[-1]
if prod_id != esa_prod_id:
scenes.remove(scene)
if len(scenes) < 1:
gs.message(_("No product found"))
return
scenes_df = pandas.DataFrame.from_dict(scenes)
if sortby:
# replace sortby keywords with USGS keywords
for idx, keyword in enumerate(sortby):
if keyword == "cloudcoverpercentage":
sortby[idx] = "cloud_cover"
# turn cloudcover to float to make it sortable
scenes_df["cloud_cover"] = pandas.to_numeric(
scenes_df["cloud_cover"])
elif keyword == "ingestiondate":
sortby[idx] = "acquisition_date"
# what does sorting by footprint mean
elif keyword == "footprint":
sortby[idx] = "display_id"
self._products_df_sorted = scenes_df.sort_values(sortby,
ascending=[asc] *
len(sortby),
ignore_index=True)
else:
self._products_df_sorted = scenes_df
gs.message(
_("{} Sentinel product(s) found").format(
len(self._products_df_sorted)))