def query_for_scenes(start_date, end_date, verbose=False, limit=None):
if 'USGS_PASSWORD' in os.environ:
api_key = retry_login(verbose=verbose)
if not api_key:
print "Failed to authenticate with USGS servers"
sys.exit(1)
full_list = []
list_offset = 0
these_scenes = 'start'
chunk_size = 500
if limit is not None and limit < chunk_size:
chunk_size = limit
if verbose:
print 'search...'
while these_scenes == 'start' or len(these_scenes) == chunk_size:
these_scenes = api.search("LANDSAT_8", "EE",
start_date=start_date, end_date=end_date,
starting_number = 1+list_offset,
max_results=chunk_size)
if verbose:
print '... %d scenes' % len(these_scenes)
full_list += these_scenes
list_offset += len(these_scenes)
if limit is not None and list_offset >= limit:
break
scene_ids = [scene['entityId'] for scene in full_list]
return scene_ids
def poll_usgs():
"""
Check whether USGS has made any new scenes available. In the case of RT scenes,
we check only a few days back. In the case of T1/T2 scenes we check 4 weeks back due to
processing latencies.
"""
api_key = api.login(os.environ['USGS_USERNAME'], os.environ['USGS_PASSWORD'], save=False)['data']
tier = os.environ['TIER']
now = datetime.now()
fmt = '%Y-%m-%d'
days_prior = 4 if tier == 'RT' else 30
start_date = (now - timedelta(days=days_prior)).strftime(fmt)
end_date = now.strftime(fmt)
# This field id represents the Collection Category
where = {
20510: tier
}
result = api.search(
'LANDSAT_8_C1', 'EE', start_date=start_date, end_date=end_date, where=where, api_key=api_key)
# Strangely, the entity id is still used to obtain a download url.
entityIds = [
scene['entityId'] for scene in result['data']['results']
]
return entityIds
def search_scenes(dataset, latitud, longitud):
# Set the Hyperion and Landsat 8 dataset
#hyperion_dataset = 'EO1_HYP_PUB'
#landsat8_dataset = 'LANDSAT_8'
# Set the EarthExplorer catalog
node = 'EE'
# Set the scene ids
scenes = api.search(dataset,
node,
lat=latitud,
lng=longitud,
distance=100,
ll=None,
ur=None,
start_date='2017-02-15',
end_date=today.strftime('%Y-%m-%d'),
where=None,
max_results=50000,
starting_number=1,
sort_order="DESC",
extended=False,
api_key=None)
scenes_list = []
for scene in scenes:
scenes_list.append(scene)
return scenes_list
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 _query_by_point(self, lon, lat, date_start, date_end):
""" Query by point using usgs api.
Args:
lon, lat (float): Coordinates.
date_start, date_end (str): The start and end date.
Returns:
query object.
"""
return api.search(
self._dataset,
self._node,
lat=lat,
lng=lon,
distance=100,
# ll={ "longitude": 108.963791 ,
# "latitude": 19.845140},
# ur={ "longitude": 110.266751 ,
# "latitude": 20.831747},
start_date=date_start,
end_date=date_end,
api_key=self.api_key,
)
def query_for_scenes(start_date, end_date, verbose=False, limit=None):
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
full_list = []
list_offset = 0
these_scenes = 'start'
chunk_size = 500
if limit is not None and limit < chunk_size:
chunk_size = limit
if verbose:
print 'search...'
while these_scenes == 'start' or len(these_scenes) == chunk_size:
these_scenes = api.search("LANDSAT_8",
"EE",
start_date=start_date,
end_date=end_date,
starting_number=1 + list_offset,
max_results=chunk_size)
if verbose:
print '... %d scenes' % len(these_scenes)
full_list += these_scenes
list_offset += len(these_scenes)
if limit is not None and list_offset >= limit:
break
scene_ids = [scene['entityId'] for scene in full_list]
return scene_ids
def search(dataset, node, aoi, start_date, end_date, longitude, latitude, distance, lower_left, upper_right, where, api_key, geojson):
node = get_node(dataset, node)
if aoi == "-":
src = click.open_file('-').readlines()
aoi = json.loads(''.join([ line.strip() for line in src ]))
bbox = map(get_bbox, aoi.get('features'))[0]
lower_left = bbox[0:2]
upper_right = bbox[2:4]
if where:
# Query the dataset fields endpoint for queryable fields
fields = api.dataset_fields(dataset, node)
def format_fieldname(s):
return ''.join(c for c in s if c.isalnum()).lower()
field_lut = { format_fieldname(field['name']): field['fieldId'] for field in fields }
where = { field_lut[format_fieldname(k)]: v for k, v in where if format_fieldname(k) in field_lut }
if lower_left:
lower_left = dict(zip(['longitude', 'latitude'], lower_left))
upper_right = dict(zip(['longitude', 'latitude'], upper_right))
data = api.search(dataset, node, lat=latitude, lng=longitude, distance=distance, ll=lower_left, ur=upper_right, start_date=start_date, end_date=end_date, where=where, api_key=api_key)
if geojson:
features = map(to_geojson_feature, data)
data = { 'type': 'FeatureCollection', 'features': features }
print(json.dumps(data))
def query_for_scenes(start_date, end_date, verbose=False, limit=None):
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
full_list = []
list_offset = 0
these_scenes = 'start'
chunk_size = 500
if limit is not None and limit < chunk_size:
chunk_size = limit
if verbose:
print 'search...'
while these_scenes == 'start' or len(these_scenes) == chunk_size:
these_scenes = api.search("LANDSAT_8", "EE",
start_date=start_date, end_date=end_date,
starting_number = 1+list_offset,
max_results=chunk_size)
if verbose:
print '... %d scenes' % len(these_scenes)
full_list += these_scenes
list_offset += len(these_scenes)
if limit is not None and list_offset >= limit:
break
scene_ids = [scene['entityId'] for scene in full_list]
return scene_ids
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 search(dataset, node, aoi, start_date, end_date, longitude, latitude,
distance, lower_left, upper_right, where, geojson, extended,
api_key):
node = get_node(dataset, node)
if aoi == "-":
src = click.open_file('-')
if not src.isatty():
lines = src.readlines()
if len(lines) > 0:
aoi = json.loads(''.join([line.strip() for line in lines]))
bbox = map(get_bbox, aoi.get('features') or [aoi])[0]
lower_left = bbox[0:2]
upper_right = bbox[2:4]
if where:
# Query the dataset fields endpoint for queryable fields
resp = api.dataset_fields(dataset, node)
def format_fieldname(s):
return ''.join(c for c in s if c.isalnum()).lower()
field_lut = {
format_fieldname(field['name']): field['fieldId']
for field in resp['data']
}
where = {
field_lut[format_fieldname(k)]: v
for k, v in where if format_fieldname(k) in field_lut
}
if lower_left:
lower_left = dict(zip(['longitude', 'latitude'], lower_left))
upper_right = dict(zip(['longitude', 'latitude'], upper_right))
result = api.search(dataset,
node,
lat=latitude,
lng=longitude,
distance=distance,
ll=lower_left,
ur=upper_right,
start_date=start_date,
end_date=end_date,
where=where,
extended=extended,
api_key=api_key)
if geojson:
result = to_geojson(result)
print(json.dumps(result))
def test_search():
expected_keys = ["totalHits", "firstRecord", "nextRecord", "results", "numberReturned", "lastRecord"]
fieldid = 20516 # TODO: compute by finding 'WRS Path' in result of api.dataset_fields("LANDSAT_8_C1", "EE")
response = api.search("LANDSAT_8_C1", "EE", start_date='20170401', end_date='20170402',
where={fieldid: '032'}, max_results=10)
assert check_root_keys(response)
return
assert len(response['data']["results"]) == 10
for item in response['data']['results']:
pr = item['displayId'].split('_')[2]
assert pr[-3:] == '032'
data = response['data']
for key in expected_keys:
assert key in data
def test_search():
expected_keys = [
"totalHits", "firstRecord", "nextRecord", "results", "numberReturned",
"lastRecord"
]
response = api.search("LANDSAT_8_C1",
"EE",
start_date='20170401',
end_date='20170402',
max_results=10)
assert check_root_keys(response)
assert len(response['data']["results"]) == 10
data = response['data']
for key in expected_keys:
assert key in data
def search(dataset, node, aoi, start_date, end_date, longitude, latitude, distance, lower_left, upper_right, where, max_results, geojson, extended, api_key):
node = get_node(dataset, node)
if aoi == "-":
src = click.open_file('-')
if not src.isatty():
lines = src.readlines()
if len(lines) > 0:
aoi = json.loads(''.join([ line.strip() for line in lines ]))
bbox = map(get_bbox, aoi.get('features') or [aoi])[0]
lower_left = bbox[0:2]
upper_right = bbox[2:4]
if where:
# Query the dataset fields endpoint for queryable fields
resp = api.dataset_fields(dataset, node)
def format_fieldname(s):
return ''.join(c for c in s if c.isalnum()).lower()
field_lut = { format_fieldname(field['name']): field['fieldId'] for field in resp['data'] }
where = { field_lut[format_fieldname(k)]: v for k, v in where if format_fieldname(k) in field_lut }
if lower_left:
lower_left = dict(zip(['longitude', 'latitude'], lower_left))
upper_right = dict(zip(['longitude', 'latitude'], upper_right))
result = api.search(
dataset, node,
lat=latitude, lng=longitude, distance=distance,
ll=lower_left, ur=upper_right,
start_date=start_date, end_date=end_date,
where=where, max_results=max_results,
extended=extended, api_key=api_key)
if geojson:
result = to_geojson(result)
print(json.dumps(result))
def querySceneLists(collection, ll, ur, start_date, end_date, api_key):
# TODO: option to order with just a point. Look back at usgs.api.search options
""" Send a request to earth explorer api
Args:
collection (string) one of 'LSR_LANDSAT_ETM_COMBINED', 'LSR_LANDSAT_8', and 'LSR_LANDSAT_TM'
ll (dict) lowerLeft corner dict with longitude and latitude keys
ur (dict) upperRight corner dict with longitude and latitude keys
dates (strings) with '%Y-%m-%d' format
api_key (string) usgs api key (retrieve it using the 'usgs login' command line)
"""
scenes = api.search(collection, 'EE',\
ll=ll,\
ur=ur,\
start_date=start_date,\
end_date=end_date,\
api_key=api_key)
scene_list = []
for scene in scenes:
scene_list.append(scene['entityId'])
return scene_list
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 retry_search(start_date, end_date, starting_number, max_results, retries=4, verbose=False):
"""
Retry USGS search multiple times, with exponential backoff between.
Required due to rate-limits imposed by USGS.
"""
if verbose:
logging.info("searching ....")
sleep_time = 5
for _ in xrange(retries + 1):
try:
scenes = api.search("LANDSAT_8", "EE",
start_date=start_date, end_date=end_date,
starting_number=starting_number,
max_results=max_results)
return scenes
except USGSError:
logging.info("USGS search failed. Retry in %s" % sleep_time)
time.sleep(sleep_time)
sleep_time *= backoff_factor(2)
return None
def test_search():
expected_keys = [
"totalHits", "firstRecord", "nextRecord", "results", "numberReturned",
"lastRecord"
]
fieldid = 20516 # TODO: compute by finding 'WRS Path' in result of api.dataset_fields("LANDSAT_8_C1", "EE")
response = api.search("LANDSAT_8_C1",
"EE",
start_date='20170401',
end_date='20170402',
where={fieldid: '032'},
max_results=10)
assert check_root_keys(response)
return
assert len(response['data']["results"]) == 10
for item in response['data']['results']:
pr = item['displayId'].split('_')[2]
assert pr[-3:] == '032'
data = response['data']
for key in expected_keys:
assert key in data
def poll_usgs():
api_key = api.login(os.environ['USGS_USERNAME'],
os.environ['USGS_PASSWORD'],
save=False)['data']
now = datetime.now()
fmt = '%Y%m%d'
start_date = (now - timedelta(days=7)).strftime(fmt)
end_date = now.strftime(fmt)
where = {
20510: 'T1' # This field id represents the Collection Category
}
result = api.search('LANDSAT_8_C1',
'EE',
start_date=start_date,
end_date=end_date,
where=where,
api_key=api_key)
# Strangely, the entity id is still used to obtain a download url.
return [scene['entityId'] for scene in result['data']['results']]
) as json_data_file: # https://martin-thoma.com/configuration-files-in-python/
authdat = json.load(json_data_file)
#Get temporary API key
usgs_api_key = api.login(str(authdat["username"]),
str(authdat["password"]),
save=False,
catalogId='EE')['data']
#Get list of scenes from Landsat 7 for 2000
lss_LC7_2000 = api.search(dataset='LANDSAT_ETM_C1',
node='EE',
ll={
"longitude": pelleextent_wgs84.XMin,
"latitude": pelleextent_wgs84.YMin
},
ur={
"longitude": pelleextent_wgs84.XMax,
"latitude": pelleextent_wgs84.YMax
},
start_date='2000-01-01',
end_date='2000-12-31',
api_key=usgs_api_key)
#Download and unzip landsat 7 scenes with least cloud cover for 2000
dl_landsattiles(lss=lss_LC7_2000,
dataset='LANDSAT_ETM_C1',
apikey=usgs_api_key,
outdir=landsatdir,
mincctile=True,
maxcloudcover=100)
def main():
API_KEY = custom_functions.get_API_key()
global no_data
_, _, _, DATA_DICT_PATH = custom_functions.get_updd()
Path_i = 1
Row_i = 1
os.makedirs(os.path.dirname(DATA_DICT_PATH), exist_ok=True)
try:
with open(DATA_DICT_PATH, 'rb') as fp:
no_data = pickle.load(fp)
try:
Path_i = sorted(no_data.keys())[-1]
Row_i = no_data[Path_i][-1]
except Exception as e:
no_data = {}
Path_i = 1
Row_i = 1
except FileNotFoundError as e:
no_data = {}
# incase path_i and row_i are taken from data_dict, they would be strings
Path_i = int(Path_i)
Row_i = int(Row_i)
if Path_i > 1 or Row_i > 1:
print(f'DATA_DICT already has data till Path {Path_i} and Row {Row_i}')
dataset = 'LANDSAT_8_C1'
for Path in range(Path_i, 234):
for Row in range(1, 249):
if (Path == Path_i) and (Row in range(1, Row_i + 1)):
continue
Path = str(Path)
if len(Path) == 1:
Path = '00' + Path
elif len(Path) == 2:
Path = '0' + Path
Row = str(Row)
if len(Row) == 1:
Row = '00' + Row
elif len(Row) == 2:
Row = '0' + Row
where = {20514: Path, 20516: Row}
try:
response = (api.search(dataset=dataset,
node='EE',
where=where,
api_key=API_KEY))
except USGSError as e:
if 'AUTH_UNAUTHORIZED' in e:
API_KEY = custom_functions.get_API_key()
response = (api.search(dataset=dataset,
node='EE',
where=where,
api_key=API_KEY))
else:
custom_functions.record_error(
e +
f'\nERROR in create_dict.py\nAPI search error at Path {Path} and Row {Row}\n'
)
continue
if (response['errorCode'] is not None):
custom_functions.record_error(
e +
f'\nERROR in create_dict.py\nerrorCode {response["errorCode"]} received during query at Path {Path} and Row {Row}\n'
)
if (response['data']['numberReturned'] == 0):
continue
no_data.setdefault(Path, [])
no_data[Path].append(Row)
print(f'Path {Path} and Row {Row} has some data')
with open(DATA_DICT_PATH, 'wb') as fp:
pickle.dump(no_data, fp)
def query(self, product_type=None, **kwargs):
"""Search for data on USGS catalogues
.. versionchanged::
1.0
* ``product_type`` is no longer mandatory
"""
product_type = kwargs.get("productType")
if product_type is None:
return [], 0
api.login(
self.config.credentials["username"],
self.config.credentials["password"],
save=True,
)
usgs_dataset = self.config.products[product_type]["dataset"]
usgs_catalog_node = self.config.products[product_type]["catalog_node"]
start_date = kwargs.pop("startTimeFromAscendingNode", None)
end_date = kwargs.pop("completionTimeFromAscendingNode", None)
footprint = kwargs.pop("geometry", None)
# Configuration to generate the download url of search results
result_summary_pattern = re.compile(
r"^ID: .+, Acquisition Date: .+, Path: (?P<path>\d+), Row: (?P<row>\d+)$" # noqa
)
# See https://pyformat.info/, on section "Padding and aligning strings" to
# understand {path:0>3} and {row:0>3}.
# It roughly means: 'if the string that will be passed as "path" has length < 3,
# prepend as much "0"s as needed to reach length 3' and same for "row"
dl_url_pattern = "{base_url}/L8/{path:0>3}/{row:0>3}/{entity}.tar.bz"
final = []
if footprint and len(footprint.keys()) == 4: # a rectangle (or bbox)
lower_left = {
"longitude": footprint["lonmin"],
"latitude": footprint["latmin"],
}
upper_right = {
"longitude": footprint["lonmax"],
"latitude": footprint["latmax"],
}
else:
lower_left, upper_right = None, None
try:
results = api.search(
usgs_dataset,
usgs_catalog_node,
start_date=start_date,
end_date=end_date,
ll=lower_left,
ur=upper_right,
)
for result in results["data"]["results"]:
r_lower_left = result["spatialFootprint"]["coordinates"][0][0]
r_upper_right = result["spatialFootprint"]["coordinates"][0][2]
summary_match = result_summary_pattern.match(
result["summary"]).groupdict()
result["geometry"] = geometry.box(r_lower_left[0],
r_lower_left[1],
r_upper_right[0],
r_upper_right[1])
# Same method as in base.py, Search.__init__()
# Prepare the metadata mapping
# Do a shallow copy, the structure is flat enough for this to be sufficient
metas = DEFAULT_METADATA_MAPPING.copy()
# Update the defaults with the mapping value. This will add any new key
# added by the provider mapping that is not in the default metadata.
# A deepcopy is done to prevent self.config.metadata_mapping from being modified when metas[metadata]
# is a list and is modified
metas.update(copy.deepcopy(self.config.metadata_mapping))
metas = mtd_cfg_as_jsonpath(metas)
result["productType"] = usgs_dataset
product_properties = properties_from_json(result, metas)
if getattr(self.config, "product_location_scheme",
"https") == "file":
product_properties["downloadLink"] = dl_url_pattern.format(
base_url="file://")
else:
product_properties["downloadLink"] = dl_url_pattern.format(
base_url=self.config.google_base_url.rstrip("/"),
entity=result["entityId"],
**summary_match)
final.append(
EOProduct(
productType=product_type,
provider=self.provider,
properties=product_properties,
geometry=footprint,
))
except USGSError as e:
logger.debug(
"Product type %s does not exist on catalogue %s",
usgs_dataset,
usgs_catalog_node,
)
logger.debug("Skipping error: %s", e)
api.logout()
return final, len(final)
from usgs import api
import json
import os
import sys
usgs_api_key = api.login(os.environ["EARTH_EXPLORER_USERNAME"],
os.environ["EARTH_EXPLORER_PASSWORD"])
scenes = \
api.search(
'EO1_HYP_PUB',
'EE',
api_key=usgs_api_key,
start_date='2017-02-01',
end_date='2017-03-01',
extended=True)
with open(sys.argv[1], "w") as f:
f.write(json.dumps(scenes))
]
import numpy as np
start_dates = [str(i) + "-06-01" for i in np.arange(2002, 2020)]
end_dates = [str(i) + "-09-30" for i in np.arange(2002, 2020)]
all_results = []
for start_date, end_date in zip(start_dates, end_dates):
results = api.search(dataset="ARD_TILE",
node="EE",
start_date=start_date,
end_date=end_date,
ll={
"longitude": -104.1064453125,
"latitude": 39.825413103424786
},
ur={
"longitude": -95.130615234375,
"latitude": 43.100982876188546
},
api_key=token)
all_results.extend(results['data']['results'])
print(start_date + " done")
filtered_results = []
for i in all_results:
if "landsat_etm_c1" in i['browseUrl']:
pass
else:
if i['cloudCover'] == None:
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
from usgs import api
import json
results = api.search('LANDSAT_TM', 'EE', start_date='2011-06-01', where={3653: 0}, max_results=1000)
print(json.dumps(results))
def geturls():
dataset = 'LANDSAT_8_C1'
global DATA_DICT_PATH
global DOWNLOAD_PATH
global PATH_S
global ROW_S
global LAST_PATH
global LAST_ROW
with open(DATA_DICT_PATH, 'rb') as fp:
data_dict = pickle.load(fp)
API_KEY = custom_functions.get_API_key()
LAST_PATH = list(data_dict.keys())[-1]
LAST_ROW = data_dict[LAST_PATH][-1]
for Path in data_dict:
# RESUME DOWNLOADING CODE
if (PATH_S != ''):
if int(Path) < int(PATH_S):
continue
for Row in data_dict[Path]:
# RESUME DOWNLOADING CODE
if (ROW_S != ''):
if int(Row) < int(ROW_S):
continue
where = {20514 : Path, 20516 : Row}
# MAKING QUERY REQUEST TO USGS API
try:
response = (api.search(dataset=dataset, node='EE', where=where, api_key=API_KEY))
except USGSError as e:
if 'AUTH_UNAUTHORIZED' in e:
API_KEY = custom_functions.get_API_key()
response = (api.search(dataset=dataset, node='EE', where=where, api_key=API_KEY))
else:
custom_functions.record_error(e + f'Error in downloader.py\nAPI search error at Path {Path} Row {Row}')
if (response['data']['numberReturned'] == 0):
continue
# FINDING OUT THE LATEST DATASET AVAILABLE FOR THE PATH-ROW COMBINATION
date = datetime.strptime('1957-10-03', '%Y-%m-%d')
for i in response['data']['results']:
currdate = datetime.strptime(i['acquisitionDate'], '%Y-%m-%d')
if(currdate > date):
date = currdate
displayId = i['displayId']
downloadUrl = i['downloadUrl']
catalog_id = response['catalog_id']
entityId = i['entityId']
yield (displayId, Path, Row, downloadUrl, catalog_id, entityId)
