def usgs_eros(self, scene, path):
""" Downloads the image from USGS """
# download from usgs if login information is provided
if self.usgs_user and self.usgs_pass:
try:
api_key = api.login(self.usgs_user, self.usgs_pass)
except USGSError as e:
error_tree = ElementTree.fromstring(str(e.message))
error_text = error_tree.find(
"SOAP-ENV:Body/SOAP-ENV:Fault/faultstring",
api.NAMESPACES).text
raise USGSInventoryAccessMissing(error_text)
download_url = api.download('LANDSAT_8',
'EE', [scene],
api_key=api_key)
if download_url:
self.output('Source: USGS EarthExplorer',
normal=True,
arrow=True)
return self.fetch(download_url[0], path)
raise RemoteFileDoesntExist(
'%s is not available on AWS S3, Google or USGS Earth Explorer'
% scene)
raise RemoteFileDoesntExist(
'%s is not available on AWS S3 or Google Storage' % scene)
def download_url(dataset, scene_ids, product, node, api_key):
node = get_node(dataset, node)
data = api.download(dataset, node, scene_ids, product)
for d in data:
click.echo(d)
def create_snapshots():
"""
Run requests against USGS API for use in tests.
"""
api_key = api.login(os.environ['USGS_USERNAME'], os.environ['USGS_PASSWORD'])
# Dataset Fields
response = api.dataset_fields("LANDSAT_8_C1", "EE", api_key=api_key)
write_response(response, 'dataset-fields.json')
# Datasets
response = api.datasets(None, "EE")
write_response(response, 'datasets.json')
# Download
response = api.download("LANDSAT_8_C1", "EE", ["LC80810712017104LGN00"], product='STANDARD')
write_response(response, 'download.json')
# Download Options
response = api.download_options("LANDSAT_8_C1", "EE", ["LC80810712017104LGN00"])
write_response(response, 'download-options.json')
# Metadata
response = api.metadata("LANDSAT_8_C1", "EE", ["LC80810712017104LGN00"])
write_response(response, 'metadata.json')
# Search
response = api.search("LANDSAT_8_C1", "EE", start_date='20170401', end_date='20170402', max_results=10)
write_response(response, 'search.json')
api.logout(api_key)
def usgs(self, scenes):
""" Downloads the image from USGS """
if not isinstance(scenes, list):
raise Exception('Expected sceneIDs list')
scene_objs = Scenes()
# download from usgs if login information is provided
if self.usgs_user and self.usgs_pass:
try:
api_key = api.login(self.usgs_user, self.usgs_pass)
except USGSError as e:
error_tree = ElementTree.fromstring(str(e.message))
error_text = error_tree.find("SOAP-ENV:Body/SOAP-ENV:Fault/faultstring", api.NAMESPACES).text
raise USGSInventoryAccessMissing(error_text)
for scene in scenes:
download_urls = api.download('LANDSAT_8', 'EE', [scene], api_key=api_key)
if download_urls:
logger.info('Source: USGS EarthExplorer')
scene_objs.add_with_files(scene, fetch(download_urls[0], self.download_dir))
else:
raise RemoteFileDoesntExist('{0} not available on AWS S3, Google or USGS Earth Explorer'.format(
' - '.join(scene)))
return scene_objs
raise RemoteFileDoesntExist('{0} not available on AWS S3 or Google Storage'.format(' - '.join(scenes)))
def fetch_dswe_images(date, ll_coord, ur_coord, output_folder, user, password,
force_login):
"""Download all DSWE images that fit the given criteria to the output folder
if they are not already present. The coordinates must be in lon/lat degrees.
"""
if not os.path.exists(output_folder):
os.mkdir(output_folder)
CATALOG = 'EE'
DATASET = 'SP_TILE_DSWE'
# Only log in if our session expired (ugly function use to check!)
if force_login or (not api._get_api_key(None)): #pylint: disable=W0212
print('Logging in to USGS EarthExplorer...')
dummy_result = api.login(user, password, save=True, catalogId=CATALOG)
print('Submitting EarthExplorer query...')
results = api.search(DATASET,
CATALOG,
where={},
start_date=date,
end_date=date,
ll=dict([('longitude', ll_coord[0]),
('latitude', ll_coord[1])]),
ur=dict([('longitude', ur_coord[0]),
('latitude', ur_coord[1])]),
max_results=12,
extended=False)
if not results['data']:
raise Exception(
'Did not find any DSWE data that matched the Landsat file!')
print('Found ' + str(len(results['data']['results'])) + ' matching files.')
for scene in results['data']['results']:
print('Found match: ' + scene['entityId'])
fname = scene['entityId'] + '.tar'
output_path = os.path.join(output_folder, fname)
if os.path.exists(output_path):
print('Already have image on disk!')
continue
r = api.download(DATASET, CATALOG, [scene['entityId']], product='DSWE')
print(r)
if not r['data']:
raise Exception('Failed to get download URL!')
url = r['data'][0]['url']
cmd = ('wget "%s" --user %s --password %s -O %s' %
(url, user, password, output_path))
print(cmd)
os.system(cmd)
if not os.path.exists(output_path):
raise Exception('Failed to download file ' + output_path)
print('Finished downloading DSWE files.')
def get_download_url(scene_root, verbose):
if 'USGS_PASSWORD' in os.environ:
if verbose:
print 'logging in...'
api_key = api.login(os.environ['USGS_USERID'],
os.environ['USGS_PASSWORD'])
if verbose:
print ' api_key = %s' % api_key
urls = api.download('LANDSAT_8', 'EE', [scene_root], 'STANDARD')
return urls[0]
def usgs_eros(self, scene, path):
""" Downloads the image from USGS """
# download from usgs if login information is provided
if self.usgs_user and self.usgs_pass:
try:
api_key = api.login(self.usgs_user, self.usgs_pass)
except USGSError as e:
error_tree = ElementTree.fromstring(str(e.message))
error_text = error_tree.find("SOAP-ENV:Body/SOAP-ENV:Fault/faultstring", api.NAMESPACES).text
raise USGSInventoryAccessMissing(error_text)
download_url = api.download('LANDSAT_8', 'EE', [scene], api_key=api_key)
if download_url:
self.output('Source: USGS EarthExplorer', normal=True, arrow=True)
return self.fetch(download_url[0], path)
raise RemoteFileDoesntExist('%s is not available on AWS S3, Google or USGS Earth Explorer' % scene)
raise RemoteFileDoesntExist('%s is not available on AWS S3 or Google Storage' % scene)
def create_snapshots():
"""
Run requests against USGS API for use in tests.
"""
api_key = api.login(os.environ['USGS_USERNAME'],
os.environ['USGS_PASSWORD'])
# Dataset Fields
response = api.dataset_fields("LANDSAT_8_C1", "EE", api_key=api_key)
write_response(response, 'dataset-fields.json')
# Datasets
response = api.datasets(None, "EE")
write_response(response, 'datasets.json')
# Download
response = api.download("LANDSAT_8_C1",
"EE", ["LC80810712017104LGN00"],
product='STANDARD')
write_response(response, 'download.json')
# Download Options
response = api.download_options("LANDSAT_8_C1", "EE",
["LC80810712017104LGN00"])
write_response(response, 'download-options.json')
# Metadata
response = api.metadata("LANDSAT_8_C1", "EE", ["LC80810712017104LGN00"])
write_response(response, 'metadata.json')
# Search
response = api.search("LANDSAT_8_C1",
"EE",
start_date='20170401',
end_date='20170402',
max_results=10)
write_response(response, 'search.json')
api.logout(api_key)
def usgs(self, scenes):
""" Downloads the image from USGS """
if not isinstance(scenes, list):
raise Exception('Expected sceneIDs list')
scene_objs = Scenes()
# download from usgs if login information is provided
if self.usgs_user and self.usgs_pass:
try:
api_key = api.login(self.usgs_user, self.usgs_pass)
except USGSError as e:
error_tree = ElementTree.fromstring(str(e.message))
error_text = error_tree.find(
"SOAP-ENV:Body/SOAP-ENV:Fault/faultstring",
api.NAMESPACES).text
raise USGSInventoryAccessMissing(error_text)
for scene in scenes:
download_urls = api.download('LANDSAT_8',
'EE', [scene],
api_key=api_key)
if download_urls:
logger.info('Source: USGS EarthExplorer')
scene_objs.add_with_files(
scene, fetch(download_urls[0], self.download_dir))
else:
raise RemoteFileDoesntExist(
'{0} not available on AWS S3, Google or USGS Earth Explorer'
.format(' - '.join(scene)))
return scene_objs
raise RemoteFileDoesntExist(
'{0} not available on AWS S3 or Google Storage'.format(
' - '.join(scenes)))
def get_download_url(scene_root, verbose):
urls = api.download('LANDSAT_8', 'EE', [scene_root], 'STANDARD')
return urls[0]
def main(argsIn): #pylint: disable=R0914,R0912
try:
usage = "usage: fetch_hdds_images.py [options]"
parser = argparse.ArgumentParser(usage=usage)
parser.add_argument("--output-folder",
dest="output_folder",
required=True,
help="Download files to this folder.")
parser.add_argument("--user",
dest="user",
required=True,
help="User name for EarthExplorer website.")
parser.add_argument("--password",
dest="password",
required=True,
help="Password name for EarthExplorer website.")
parser.add_argument(
"--force-login",
action="store_true",
dest="force_login",
default=False,
help="Don't reuse the cached EE API key if present.")
parser.add_argument("--refetch-datasets",
action="store_true",
dest="refetch_datasets",
default=False,
help="Force a refetch of the dataset list.")
parser.add_argument(
"--refetch-scenes",
action="store_true",
dest="refetch_scenes",
default=False,
help="Force refetches of scene lists for each dataset.")
parser.add_argument(
"--image-list-path",
dest="image_list_path",
default=None,
help=
"Path to text file containing list of image IDs to download, one per line."
)
parser.add_argument("--event-name",
dest="event_name",
default=None,
help="Only download images from this event.")
options = parser.parse_args(argsIn)
except argparse.ArgumentError:
print(usage)
return -1
if options.output_folder and not os.path.exists(options.output_folder):
os.mkdir(options.output_folder)
images_to_use = []
if options.image_list_path:
with open(options.image_list_path, 'r') as f:
for line in f:
images_to_use.append(line.strip())
# Only log in if our session expired (ugly function use to check!)
if options.force_login or (not api._get_api_key(None)): #pylint: disable=W0212
print('Logging in to USGS EarthExplorer...')
api.login(options.user, options.password, save=True, catalogId=CATALOG) #pylint: disable=W0612
print(api._get_api_key(None)) #pylint: disable=W0212
raise Exception('DEBUG')
# Retrieve all of the available datasets
dataset_list = get_dataset_list(options)
print('Found ' + str(len(dataset_list)) + ' useable datasets.')
#raise Exception('debug')
# Don't think we need to do this!
#get_dataset_fields(dataset_list)
# TODO: Work through some errors.
counter = 0
for (dataset, full_name) in dataset_list:
counter = counter + 1
#if counter == 1:
# continue
if options.event_name: # Only download images from the specified event
if options.event_name.lower() not in full_name.lower():
continue
dataset_folder = os.path.join(options.output_folder, full_name)
scene_list_path = os.path.join(dataset_folder, 'scene_list.dat')
done_flag_path = os.path.join(dataset_folder, 'done.flag')
if not os.path.exists(dataset_folder):
os.mkdir(dataset_folder)
if os.path.exists(done_flag_path) and not options.refetch_scenes:
print('Skipping completed dataset ' + full_name)
continue
print('--> Search scenes for: ' + full_name)
BATCH_SIZE = 10000
if not os.path.exists(scene_list_path) or options.refetch_scenes:
# Request the scene list from USGS
#details = {'Agency - Platform - Vendor':'WORLDVIEW', 'Sensor Type':'MS'}
#details = {'sensor_type':'MS'}
details = {} # TODO: How do these work??
# Large sets of results require multiple queries in order to get all of the data
done = False
error = False
all_scenes = [] # Acculumate all scene data here
while not done:
print('Searching with start offset = ' + str(len(all_scenes)))
results = api.search(dataset,
CATALOG,
where=details,
max_results=BATCH_SIZE,
starting_number=len(all_scenes),
extended=False)
if 'results' not in results['data']:
print('ERROR: Failed to get any results for dataset: ' +
full_name)
error = True
break
if len(results['data']['results']) < BATCH_SIZE:
done = True
all_scenes += results['data']['results']
if error:
continue
results['data']['results'] = all_scenes
# Cache the results to disk
with open(scene_list_path, 'wb') as f:
pickle.dump(results, f)
else: # Load the results from the cache file
with open(scene_list_path, 'rb') as f:
results = pickle.load(f)
print('Got ' + str(len(results['data']['results'])) +
' scene results.')
for scene in results['data']['results']:
fail = False
REQUIRED_PARTS = ['displayId', 'summary', 'entityId', 'displayId']
for p in REQUIRED_PARTS:
if (p not in scene) or (not scene[p]):
print('scene object is missing element: ' + p)
print(scene)
fail = True
if fail:
continue
# If image list was provided skip other image names
if images_to_use and (scene['displayId'] not in images_to_use):
continue
# Figure out the downloaded file path for this image
file_name = scene['displayId'] + '.zip'
output_path = os.path.join(dataset_folder, file_name)
if not os.path.exists(dataset_folder):
os.mkdir(dataset_folder)
if os.path.exists(output_path):
continue # Already have the file!
# Check if this is one of the sensors we are interested in.
DESIRED_SENSORS = [('worldview', 'hp'),
('worldview', 'msi')] # TODO: Add more
parts = scene['summary'].lower().split(',')
platform = None
sensor = None
for part in parts:
if 'platform:' in part:
platform = part.split(':')[1].strip()
if 'sensor:' in part:
sensor = part.split(':')[1].strip()
if (not platform) or (not sensor):
raise Exception('Unknown sensor: ' + scene['summary'])
if (platform, sensor) not in DESIRED_SENSORS:
print((platform, sensor))
print('Undesired sensor: ' + scene['summary'])
continue
# Investigate the number of bands
PLATFORM_BAND_COUNTS = {'worldview': 8, 'TODO': 1}
min_num_bands = PLATFORM_BAND_COUNTS[platform]
num_bands = None
try:
meta = api.metadata(dataset, CATALOG, scene['entityId'])
except json.decoder.JSONDecodeError:
print('Error fetching metadata for dataset = ' + dataset +
', entity = ' + scene['entityId'])
continue
try:
for m in meta['data'][0]['metadataFields']:
if m['fieldName'] == 'Number of bands':
num_bands = int(m['value'])
break
if not num_bands:
raise KeyError() # Treat like the except case
if num_bands < min_num_bands:
print('Skipping %s, too few bands: %d' %
(scene['displayId'], num_bands))
continue
except KeyError:
print('Unable to perform metadata check!')
print(meta)
# Make sure we know which file option to download
try:
types = api.download_options(dataset, CATALOG,
scene['entityId'])
except json.decoder.JSONDecodeError:
print('Error decoding download options!')
continue
if not types['data'] or not types['data'][0]:
raise Exception('Need to handle types: ' + str(types))
ready = False
download_type = 'STANDARD' # TODO: Does this ever change?
for o in types['data'][0]['downloadOptions']:
if o['available'] and o['downloadCode'] == download_type:
ready = True
break
if not ready:
raise Exception('Missing download option for scene: ' +
str(types))
# Get the download URL of the file we want.
r = api.download(dataset,
CATALOG, [scene['entityId']],
product=download_type)
try:
url = r['data'][0]['url']
except Exception as e:
raise Exception('Failed to get download URL from result: ' +
str(r)) from e
print(scene['summary'])
# Finally download the data!
cmd = ('wget "%s" --user %s --password %s -O %s' %
(url, options.user, options.password, output_path))
print(cmd)
os.system(cmd)
#raise Exception('DEBUG')
print('Finished processing dataset: ' + full_name)
os.system('touch ' + done_flag_path) # Mark this dataset as finished
#raise Exception('DEBUG')
#if not os.path.exists(output_path):
# raise Exception('Failed to download file ' + output_path)
print('Finished downloading HDDS! files.')
# Can just let this time out
#api.logout()
return 0
def download_url(dataset, scene_ids, product, node, api_key):
node = get_node(dataset, node)
data = api.download(dataset, node, scene_ids, product)
print(json.dumps(data))
def test_download():
response = api.download("LANDSAT_8_C1",
"EE", ["LC80810712017104LGN00"],
product='STANDARD')
assert check_root_keys(response)
assert len(response['data']) == 1
def get_download_url(scene_root, verbose):
urls = api.download('LANDSAT_8', 'EE', [scene_root], 'STANDARD')
return urls[0]
def test_download():
response = api.download("LANDSAT_8_C1", "EE", ["LC80810712017104LGN00"], product='STANDARD')
assert check_root_keys(response)
assert len(response['data']) == 1
def download_url(dataset, scene_ids, product, node, api_key):
node = get_node(dataset, node)
data = api.download(dataset, node, scene_ids, product)
click.echo(json.dumps(data))
def dl_landsattiles(lss,
dataset,
apikey,
outdir,
mincctile=True,
maxcloudcover=100):
sublss = defaultdict(dict)
print('Subsetting scenes to keep those with minimum land cover...')
if mincctile:
for tile in lss['data']['results']:
rowpath = tile['displayId'].split('_')[2]
if rowpath in sublss:
if float(tile['cloudCover']) < float(
sublss[rowpath]['cloudCover']):
sublss[rowpath] = tile
else:
sublss[rowpath] = tile
else:
sublss = lss
for tile in sublss.values():
# Only continue if there is less than set cloud cover on the image
if float(tile['cloudCover']) < maxcloudcover:
print(tile['entityId'])
# Get tile download info
gettile = api.download(dataset=dataset,
node='EE',
entityids=tile['entityId'],
product='STANDARD',
api_key=apikey)
tileurl = gettile['data'][0]['url']
# Get file name
print('Trying to access {}...'.format(tileurl))
req = urlopen_with_retry(in_url=tileurl)
cd = req.headers.get('content-disposition')
fname = re.findall('(?<=filename=")[A-Za-z0-9_.]+', cd)[0]
tiledirname = os.path.join(
outdir,
os.path.splitext(os.path.splitext(fname)[0])[0])
if not os.path.exists(tiledirname):
os.mkdir(tiledirname)
outpath = os.path.join(tiledirname, fname)
if not os.path.exists(
outpath
): # Inspired from https://github.com/yannforget/landsatxplore
with open(outpath, 'wb') as f:
chunk = req.read()
f.write(chunk)
print('Downloaded {}, now unzipping...'.format(outpath))
tarf = os.path.splitext(outpath)[0]
with gzip.GzipFile(outpath, 'rb') as input:
s = input.read()
with open(tarf, 'wb') as output:
output.write(s)
with tarfile.open(tarf) as ftar:
ftar.extractall(
tiledirname) # specify which folder to extract to
else:
print('{} already exists...'.format(outpath))
