def get_user_info(self, user_id):
""" GET to CKAN API to get list of admins
https://docs.ckan.org/en/2.8/api/#ckan.logic.action.get.user_show
"""
url = '{}{}?id={}'.format(self.base_url, self.user_show_url, user_id)
headers = self.get_request_headers(include_api_key=True)
logger.info(f'GET {url} headers:{headers}')
try:
req = requests.get(url, headers=headers)
except Exception as e:
error = 'ERROR getting users information: {} [{}]'.format(url, e)
raise
content = req.content
if req.status_code >= 400:
error = 'ERROR getting users information: {} \n\t Status code: {} \n\t content:{}'.format(
url, req.status_code, content)
logger.error(error)
raise Exception(error)
try:
json_content = json.loads(content)
except Exception as e:
error = 'ERROR parsing JSON data from users information {} [{}]'.format(
content, e)
raise
if not json_content['success']:
error = 'API response failed: {}'.format(
json_content.get('error', None))
logger.error(error)
return json_content
def validate_origin_dataset(self):
# check required https://docs.ckan.org/en/2.8/api/#ckan.logic.action.create.package_create
if self.ckan_owner_org_id is None:
error = 'Owner organization ID is required'
self.errors.append(error)
return False
requireds = []
if self.schema == 'usmetadata':
requireds += ['accessLevel', 'identifier',
'contactPoint__fn', 'programCode',
'bureauCode', 'contactPoint__hasEmail',
'publisher', 'modified', 'keyword']
ok = True
for req in requireds:
# read fields considering the __ separator
identified = self.identify_origin_element(raw_field=req)
if identified in [None, '']:
error = f'"{req}" field could not be empty at origin dataset'
self.errors.append(error)
ok = False
if not ok:
logger.info(f'requires failed on {self.original_dataset}: {self.errors}')
return ok
def show_package(self, ckan_package_id_or_name):
""" GET to CKAN API to show a package/dataset """
url = '{}{}'.format(self.base_url, self.package_show_url)
headers = self.get_request_headers(include_api_key=True)
data = {'id': ckan_package_id_or_name}
logger.info(f'GET {url} headers:{headers} data:{data}')
try:
req = requests.get(url, params=data, headers=headers)
except Exception as e:
error = 'ERROR showing CKAN package: {} [{}]'.format(url, e)
raise
content = req.content
if req.status_code >= 400:
error = 'ERROR showing CKAN package: {} \n\t Status code: {} \n\t content:{}'.format(
url, req.status_code, content)
logger.error(error)
raise Exception(error)
content = req.content
try:
json_content = json.loads(content)
except Exception as e:
error = 'ERROR parsing JSON data from show_package: {} [{}]'.format(
content, e)
raise
if not json_content['success']:
error = 'API response failed: {}'.format(
json_content.get('error', None))
logger.error(error)
return json_content
def transform_to_ckan_dataset(self, existing_resources=None):
valid, error = self.validate_origin_dataset()
if not valid:
raise Exception(f'Error validating origin dataset: {error}')
dataset = self.original_dataset.get('iso_values', {})
tags = dataset.get('tags', [])
cleaned_tags = clean_tags(tags)
self.ckan_dataset['tag_string'] = ','.join(cleaned_tags)
# previous transformations at origin
for old_field, field_ckan in self.mapped_fields.items():
logger.debug(f'Connecting fields "{old_field}", "{field_ckan}"')
# identify origin and set value to destination
origin = self.identify_origin_element(raw_field=old_field)
if origin is None:
logger.debug(f'No data in origin for "{old_field}"')
else:
self.set_destination_element(raw_field=field_ckan, new_value=origin)
logger.debug(f'Connected OK fields "{old_field}"="{origin}"')
self.infer_resources()
self.ckan_dataset['resources'] = self.transform_resources()
# custom changes
self.fix_licence_url()
self.set_browse_graphic()
self.set_temporal_extent()
self.set_responsible_party()
self.set_bbox()
# define name (are uniques in CKAN instance)
if 'name' not in self.ckan_dataset or self.ckan_dataset['name'] == '':
self.ckan_dataset['name'] = self.generate_name(title=self.ckan_dataset['title'])
# mandatory
self.ckan_dataset['owner_org'] = self.ckan_owner_org_id
# clean all empty unused values (can't pop keys while iterating)
ckan_dataset_copy = self.ckan_dataset.copy()
for k, v in self.ckan_dataset.items():
if v is None:
ckan_dataset_copy.pop(k)
self.ckan_dataset = ckan_dataset_copy
valid = self.validate_final_dataset()
if not valid:
raise Exception(f'Error validating final dataset: {self.errors} from {self.original_dataset}')
logger.info('Dataset transformed {} OK'.format(self.original_dataset.get('identifier', '')))
return self.ckan_dataset
def get_xml_tree(self):
if self.xml_tree is None:
parser = letree.XMLParser(remove_blank_text=True)
if type(self.xml_str) != str:
logger.info('XML_STR is not str, is {}: {}'.format(
type(self.xml_str), self.xml_str))
xml_str = str(self.xml_str)
else:
xml_str = self.xml_str
# logger.debug(f'Parsing ISO XML {xml_str}')
self.xml_tree = letree.fromstring(xml_str, parser=parser)
return self.xml_tree
def delete_all_harvest_sources(self,
harvest_type='harvest',
source_type='datajson'):
logger.info(f'Deleting local harvest sources from {self.base_url}')
deleted = []
for harvest_sources in self.search_harvest_packages(
harvest_type=harvest_type, source_type=source_type):
for harvest_source in harvest_sources:
harvest_source_name = harvest_source['name']
if harvest_source_name in deleted:
#TODO fix duplicated
continue
logger.info(f'Deleting local harvest {harvest_source_name}')
res = self.delete_package(
ckan_package_id_or_name=harvest_source_name)
if not res['success']:
raise Exception(f'Failed to delete {harvest_source_name}')
else:
logger.info(f'Deleted {harvest_source_name}')
deleted.append(harvest_source_name)
deleted += 1
logger.info(f'{deleted} harvest sources deleted')
return deleted
def create_organization(self, organization, check_if_exists=True):
""" POST to CKAN API to create a new organization
organization is just a python dict
https://docs.ckan.org/en/2.8/api/#ckan.logic.action.create.organization_create
"""
logger.info(f'**** Creating Organization {organization}')
if check_if_exists:
logger.info(f'Exists Organization? {organization}')
res = self.show_organization(
organization_id_or_name=organization['name'])
if res['success']:
# do not create
logger.info(f'Avoid create Organization {organization}')
return res
url = '{}{}'.format(self.base_url, self.organization_create_url)
headers = self.get_request_headers(include_api_key=True)
headers['Content-Type'] = 'application/json'
organization = json.dumps(organization)
logger.info(f'POST {url} headers:{headers} data:{organization}')
try:
req = requests.post(url, data=organization, headers=headers)
except Exception as e:
error = 'ERROR creating [POST] organization: {} [{}]'.format(
url, e)
raise
content = req.content
if req.status_code >= 400:
error = ('ERROR creating [STATUS] organization: {}'
'\n\t Status code: {}'
'\n\t content:{}'
'\n\t Dataset {}'.format(url, req.status_code, content,
organization))
logger.error(error)
raise Exception(error)
try:
json_content = json.loads(content)
except Exception as e:
error = 'ERROR parsing JSON data: {} [{}]'.format(content, e)
logger.error(error)
raise
if not json_content['success']:
error = 'API response failed: {}'.format(
json_content.get('error', None))
logger.error(error)
return json_content
def import_harvest_sources(
self,
catalog_url,
method='GET', # depend on CKAN version, GET for older versions
on_duplicated='DELETE',
harvest_type='harvest',
source_type='datajson',
delete_local_harvest_sources=True):
""" import harvest sources from another CKAN open data portal """
if delete_local_harvest_sources:
deleted = self.delete_all_harvest_sources(source_type=source_type)
logger.info(f'Getting external harvest sources for {catalog_url}')
external_portal = CKANPortalAPI(base_url=catalog_url)
total_sources = 0
search_external = external_portal.search_harvest_packages(
method=method, harvest_type=harvest_type, source_type=source_type)
for external_harvest_sources in search_external:
for external_harvest_source in external_harvest_sources:
name = external_harvest_source['name']
organization = external_harvest_source['organization']
logger.info(f'**** Importing Organization {organization}')
# copy organization locally
del organization['id'] # drop original ID
del organization['created']
del organization['revision_id']
res = self.create_organization(organization=organization)
owner_org_id = organization['name']
config = external_harvest_source.get('config', {})
# res = self.delete_package(name)
logger.info(external_harvest_source)
res = self.create_harvest_source(
title=external_harvest_source['title'],
url=external_harvest_source['url'],
owner_org_id=owner_org_id,
name=name,
config=config,
notes=external_harvest_source['notes'],
source_type=source_type,
frequency=external_harvest_source['frequency'],
on_duplicated=on_duplicated)
if not res['success']:
raise Exception(f'Failed to import harvest source {name}')
else:
logger.info(f'Created {name}')
total_sources += 1
return total_sources
def fetch(self, timeout=30):
""" download de data.json file """
logger.info(f'Fetching data from {self.url}')
if self.url is None:
error = "No URL defined"
self.errors.append(error)
logger.error(error)
raise Exception(error)
try:
req = requests.get(self.url, timeout=timeout)
except Exception as e:
error = 'ERROR Donwloading data: {} [{}]'.format(self.url, e)
self.errors.append(error)
logger.error(error)
raise
logger.info(f'Data fetched status {req.status_code}')
if req.status_code >= 400:
error = '{} HTTP error: {}'.format(self.url, req.status_code)
self.errors.append(error)
logger.error(error)
raise Exception(error)
logger.info(f'Data fetched OK')
self.raw_data_json = req.content
def show_organization(
self,
organization_id_or_name,
method='POST'): # troubles using 2.3 and 2.8 CKAN versions):
""" GET to CKAN API to show a organization """
url = '{}{}'.format(self.base_url, self.organization_show_url)
headers = self.get_request_headers()
data = {'id': organization_id_or_name}
logger.info(f'POST {url} headers:{headers} data:{data}')
try:
if method == 'POST':
req = requests.post(url, data=data, headers=headers)
else:
req = requests.get(url, params=data, headers=headers)
except Exception as e:
error = 'ERROR showing organization: {} [{}]'.format(url, e)
raise
content = req.content
if req.status_code >= 400 and req.status_code != 404:
error = 'ERROR showing organization: {} \n\t Status code: {} \n\t content:{}'.format(
url, req.status_code, content)
logger.error(error)
raise Exception(error)
try:
json_content = json.loads(content)
except Exception as e:
error = 'ERROR parsing JSON data from show_organization: {} [{}]'.format(
content, e)
raise
if not json_content['success']:
error = 'API response failed: {}'.format(
json_content.get('error', None))
logger.error(error)
return json_content
def update_package(self, ckan_package):
""" POST to CKAN API to update a package/dataset
ckan_package is just a python dict
https://docs.ckan.org/en/2.8/api/#ckan.logic.action.update.package_update
"""
url = '{}{}'.format(self.base_url, self.package_update_url)
headers = self.get_request_headers(include_api_key=True)
headers['Content-Type'] = 'application/json'
ckan_package = json.dumps(ckan_package)
logger.info(f'POST {url} headers:{headers} data:{ckan_package}')
try:
req = requests.post(url, data=ckan_package, headers=headers)
except Exception as e:
error = 'ERROR creating CKAN package: {} [{}]'.format(url, e)
raise
content = req.content
if req.status_code >= 400:
error = 'ERROR updateing CKAN package: {} \n\t Status code: {} \n\t content:{}'.format(
url, req.status_code, content)
logger.error(error)
raise Exception(error)
try:
json_content = json.loads(content)
except Exception as e:
error = 'ERROR parsing JSON data: {} [{}]'.format(content, e)
raise
if not json_content['success']:
error = 'API response failed: {}'.format(
json_content.get('error', None))
logger.error(error)
return json_content
def transform_to_ckan_dataset(self, existing_resources=None):
# check how to parse
# https://github.com/GSA/ckanext-datajson/blob/07ca20e0b6dc1898f4ca034c1e073e0c27de2015/ckanext/datajson/parse_datajson.py#L5
# if we are updating existing dataset we need to merge resources
logger.info('Transforming data.json dataset {}'.format(self.original_dataset.get('identifier', '')))
valid = self.validate_origin_dataset()
if not valid:
# raise Exception(f'Error validating origin dataset: {error}')
return None
datajson_dataset = self.original_dataset
tags = datajson_dataset.get('keyword', [])
cleaned_tags = clean_tags(tags)
self.ckan_dataset['tag_string'] = ','.join(cleaned_tags)
# previous transformations at origin
for old_field, field_ckan in self.mapped_fields.items():
logger.debug(f'Connecting fields "{old_field}", "{field_ckan}"')
# identify origin and set value to destination
origin = self.identify_origin_element(raw_field=old_field)
if origin is None:
logger.debug(f'No data in origin for "{old_field}"')
else:
self.set_destination_element(raw_field=field_ckan, new_value=origin)
logger.debug(f'Connected OK fields "{old_field}"="{origin}"')
# transform distribution into resources
distribution = datajson_dataset['distribution'] if 'distribution' in datajson_dataset else []
# if _distribution_ is empty then we try to create them from "accessURL" or "webService" URLs
if distribution is None or distribution == []:
distribution = self.infer_resources()
self.ckan_dataset['resources'] = self.transform_resources(distribution)
# move out the resources with validation errores
# and log the error as a dataset error
final_resources = []
for resource in self.ckan_dataset['resources']:
if 'error' in resource:
self.errors.append(resource)
else:
final_resources.append(resource)
self.ckan_dataset['resources'] = final_resources
if existing_resources is not None:
res = self.merge_resources(existing_resources=existing_resources, new_resources=self.ckan_dataset['resources'])
self.ckan_dataset['resources'] = res
# add custom extras
# add source_datajson_identifier = {"key": "source_datajson_identifier", "value": True}
self.set_destination_element(raw_field='extras__source_datajson_identifier', new_value=True)
# define name (are uniques in CKAN instance)
if 'name' not in self.ckan_dataset or self.ckan_dataset['name'] == '':
name = self.generate_name(title=self.ckan_dataset['title'])
self.ckan_dataset['name'] = name
# mandatory
self.ckan_dataset['owner_org'] = self.ckan_owner_org_id
# check for license
if datajson_dataset.get('license', None) not in [None, '']:
original_license = datajson_dataset['license']
original_license = original_license.replace('http://', '')
original_license = original_license.replace('https://', '')
original_license = original_license.rstrip('/')
license_id = ckan_settings.LICENCES.get(original_license, "other-license-specified")
self.ckan_dataset['license_id'] = license_id
# define publisher as extras as we expect
publisher = datajson_dataset.get('publisher', None)
if publisher is not None:
publisher_name = publisher.get('name', '')
# TODO check which place we are going to use
self.set_extra('publisher', publisher_name)
# self.ckan_dataset['publisher'] = publisher_name
parent_publisher = publisher.get('subOrganizationOf', None)
if parent_publisher is not None:
publisher_hierarchy = [publisher_name]
while parent_publisher:
parent_name = parent_publisher.get('name', '')
parent_publisher = parent_publisher.get('subOrganizationOf', None)
publisher_hierarchy.append(parent_name)
publisher_hierarchy.reverse()
publisher_hierarchy = " > ".join(publisher_hierarchy)
self.set_extra('publisher_hierarchy', publisher_hierarchy)
# clean all empty unused values (can't pop keys while iterating)
ckan_dataset_copy = self.ckan_dataset.copy()
for k, v in self.ckan_dataset.items():
if v is None:
ckan_dataset_copy.pop(k)
self.ckan_dataset = ckan_dataset_copy
valid = self.validate_final_dataset()
if valid is None:
return None
logger.info('Dataset transformed {} OK'.format(self.original_dataset.get('identifier', '')))
return ckan_dataset_copy
def test_create_harvest_source(self):
logger.info(f'Creating harvest source from {CKAN_BASE_URL}')
cpa = CKANPortalAPI(base_url=CKAN_BASE_URL, api_key=CKAN_API_KEY)
try:
cpa.delete_all_harvest_sources(harvest_type='harvest',
source_type='datajson')
except Exception as e:
logger.error(f'Error cleaning previous harvest soures {e}')
pass
title = 'Energy JSON test {}'.format(random.randint(1, 999999))
url = 'http://www.energy.gov/data-{}.json'.format(
random.randint(1, 999999))
res = cpa.create_harvest_source(
title=title,
url=url,
owner_org_id=CKAN_ORG_ID,
source_type='datajson',
notes='Some tests about local harvesting sources creation',
frequency='WEEKLY')
self.assertTrue(res['success'])
harvest_source = res['result']
logger.info('Created: {}'.format(res['success']))
# read it
res = cpa.show_package(ckan_package_id_or_name=harvest_source['id'])
self.assertTrue(res['success'])
self.assertEqual(harvest_source['url'], url)
self.assertEqual(harvest_source['title'], title)
self.assertEqual(harvest_source['type'], 'harvest')
self.assertEqual(harvest_source['source_type'], 'datajson')
# search for it
results = cpa.search_harvest_packages(rows=1000,
harvest_type='harvest',
source_type='datajson')
created_ok = False
for datasets in results:
for dataset in datasets:
# print('FOUND: {}'.format(dataset['name']))
if dataset['name'] == harvest_source['name']:
created_ok = True
logger.info('Found!')
else:
logger.info('Other harvest source: {}'.format(
dataset['name']))
assert created_ok == True
# create a dataset with this harvest_soure_id
dataset_title = 'Dataset number {}'.format(random.randint(1, 999999))
dataset_name = slugify(dataset_title)
tags = [{'name': 'tag81'}, {'name': 'tag82'}]
randval = random.randint(1, 999)
extras = [
{
'key': 'harvest_source_id',
'value': harvest_source['id']
},
{
'key': 'harvest_source_title',
'value': harvest_source['title']
},
# {'key': 'harvest_object_id', 'value': harvest_source['id']}, # ? not sure
{
'key': 'harvest_ng_source_id',
'value': harvest_source['id']
},
{
'key': 'harvest_ng_source_title',
'value': harvest_source['title']
},
{
'key': 'try_a_extra',
'value': randval
}
]
package = {
'name': dataset_name,
'title': dataset_title,
'owner_org': CKAN_ORG_ID,
'tags': tags,
'extras': extras
}
res2 = cpa.create_package(ckan_package=package)
self.assertTrue(res2['success'])
logger.info('Package with harvest source: {}'.format(res2['success']))
# read full dataset
res3 = cpa.show_package(ckan_package_id_or_name=dataset_name)
self.assertTrue(res3['success'])
ckan_dataset = res3['result']
logger.info(
'Package with harvest source readed: {}'.format(ckan_dataset))
assert 'extras' in ckan_dataset
assert [str(randval)] == [
extra['value'] for extra in ckan_dataset['extras']
if extra['key'] == 'try_a_extra'
]
# my custom ID (not connected to a real harvest ID)
assert [harvest_source['id']] == [
extra['value'] for extra in ckan_dataset['extras']
if extra['key'] == 'harvest_ng_source_id'
]
# check if this package is related to harvest source
total_datasets_in_source = 0
datasets_from_source = cpa.search_harvest_packages(
harvest_source_id=harvest_source['id'])
connected_ok = False
for datasets in datasets_from_source:
for dataset in datasets:
total_datasets_in_source += 1
if dataset['name'] == dataset_name:
connected_ok = True
logger.info('Found!')
else:
# we just expect one dataset
error = '{} != {} ------ {}'.format(
dataset['name'], dataset_name, dataset)
logger.error(error)
assert error == False
assert connected_ok == True
assert total_datasets_in_source == 1
logger.info(
f' +++++++++++++ total_datasets_in_source={total_datasets_in_source}'
)
# this fails, harvest process is more complex that just add an extra
# assert [harvest_source['id']] == [extra['value'] for extra in ckan_dataset['extras'] if extra['key'] == 'harvest_source_id']
# delete both
logger.info('Delete CKAN package: {}'.format(ckan_dataset['id']))
res4 = cpa.delete_package(ckan_package_id_or_name=ckan_dataset['id'])
self.assertTrue(res4['success'])
logger.info('Delete Harvest source: {}'.format(harvest_source['id']))
res5 = cpa.delete_package(ckan_package_id_or_name=harvest_source['id'])
self.assertTrue(res5['success'])
def search_harvest_packages(
self,
rows=1000,
method='POST', # POST work in CKAN 2.8, fails in 2.3
harvest_source_id=None, # just one harvest source
harvest_type=None, # harvest for harvest sources
source_type=None):
""" search harvested packages or harvest sources
"rows" is the page size.
You could search for an specific harvest_source_id """
start = 0
sort = "metadata_modified desc"
url = '{}{}'.format(self.base_url, self.package_search_url)
page = 0
# TODO check for a real paginated version
while url:
page += 1
params = {'start': start, 'rows': rows} # , 'sort': sort}
if harvest_source_id is not None:
# our new extra is working
params['fq'] = f'+harvest_ng_source_id:"{harvest_source_id}"'
elif harvest_type is not None:
# at my local instance I need this.
# I not sure why, in another public instances is not needed
params['fq'] = f'+dataset_type:{harvest_type}'
if source_type is not None:
params[
'q'] = f'(type:{harvest_type} source_type:{source_type})'
else:
params['q'] = f'(type:{harvest_type})'
logger.info(
f'Searching {url} PAGE:{page} start:{start}, rows:{rows} with params: {params}'
)
headers = self.get_request_headers()
try:
logger.info(f'Search harvest packages via {method}')
if method == 'POST': # depend on CKAN version
req = requests.post(url, data=params, headers=headers)
else:
req = requests.get(url, params=params, headers=headers)
except Exception as e:
error = 'ERROR Donwloading package list: {} [{}]'.format(
url, e)
raise ValueError(
'Failed to get package list at {}'.format(url))
content = req.content
if req.status_code >= 400:
error = ('ERROR searching CKAN package: {}'
'\n\t Status code: {}'
'\n\t Params: {}'
'\n\t content:{}'.format(url, req.status_code, params,
content))
logger.error(error)
raise Exception(error)
try:
json_content = json.loads(content) # check for encoding errors
except Exception as e:
error = 'ERROR parsing JSON data: {} [{}]'.format(content, e)
raise ValueError(error)
if not json_content['success']:
error = 'API response failed: {}'.format(
json_content.get('error', None))
raise ValueError(error)
result = json_content['result']
count_results = result['count']
sort_results = result['sort']
facet_results = result['facets']
results = result['results']
real_results_count = len(results)
self.total_packages += real_results_count
logger.info(f'{real_results_count} results')
if real_results_count == 0:
url = None
else:
start += rows
self.package_list += results
yield (results)
def create_package(
self,
ckan_package,
on_duplicated='RAISE', # if name already exists 'RAISE' 'SKIP' | 'DELETE'
):
""" POST to CKAN API to create a new package/dataset
ckan_package is just a python dict
https://docs.ckan.org/en/2.8/api/#ckan.logic.action.create.package_create
Params:
- ckan_package: a dict with with a ready-to-save package
- on_duplicated (str): action to take where the package already exists:
+ RAISE: raise an error
+ SKIP: returns show_package results
+ DELETE: remove the package and try to create again
"""
url = '{}{}'.format(self.base_url, self.package_create_url)
headers = self.get_request_headers(include_api_key=True)
headers['Content-Type'] = 'application/json'
ckan_package_str = json.dumps(ckan_package)
logger.info(f'POST {url} headers:{headers} data:{ckan_package}')
try:
req = requests.post(url, data=ckan_package_str, headers=headers)
except Exception as e:
error = 'ERROR creating [POST] CKAN package: {} [{}]'.format(
url, e)
raise
content = req.content
try:
json_content = json.loads(content)
except Exception as e:
error = 'ERROR parsing JSON data: {} [{}]'.format(content, e)
logger.error(error)
raise
if req.status_code == 409:
logger.info(f'409 json_content: {json_content}')
# another posible [error] = {'owner_org': ['Organization does not exist']}
# Check for duplicates
name_errors = json_content['error'][
'url'] if 'name' in json_content['error'] else []
dataset_exists = len([
ne for ne in name_errors if "That URL is already in use" in ne
]) > 0
url_errors = json_content['error']['url'] if 'url' in json_content[
'error'] else []
harvest_exists = len([
ue for ue in url_errors
if "There already is a Harvest Source for this URL" in ue
]) > 0
is_duplicated = dataset_exists or harvest_exists
if is_duplicated:
logger.error(f'Already exists! ACTION: {on_duplicated}')
if on_duplicated == 'SKIP':
# returns {'success': True, 'result': {the package}}
res = self.show_package(
ckan_package_id_or_name=ckan_package['name'])
logger.info(f'Skipped: {res}')
return res
elif on_duplicated == 'DELETE':
delr = self.delete_package(
ckan_package_id_or_name=ckan_package['name'])
if not delr['success']:
raise Exception('Failed to delete {}'.format(
ckan_package['name']))
return self.create_package(ckan_package=ckan_package,
on_duplicated='RAISE')
elif on_duplicated == 'RAISE':
error = ('DUPLICATED CKAN package: {}'
'\n\t Status code: {}'
'\n\t content:{}'
'\n\t Dataset {}'.format(url, req.status_code,
content, ckan_package))
logger.error(error)
raise Exception(error)
if req.status_code >= 400:
error = ('ERROR creating CKAN package: {}'
'\n\t Status code: {}'
'\n\t content:{}'
'\n\t Dataset {}'.format(url, req.status_code, content,
ckan_package))
logger.error(error)
raise Exception(error)
if not json_content['success']:
error = 'API response failed: {}'.format(
json_content.get('error', None))
logger.error(error)
logger.info(f'Harvest source created: {json_content}')
return json_content
def search_packages(
self,
rows=1000,
method='POST', # POST work in CKAN 2.8, fails in 2.3
search_params={}): # datajson for
""" search packages.
"rows" is the page size.
"""
start = 0
url = '{}{}'.format(self.base_url, self.package_search_url)
page = 0
# TODO check for a real paginated version
while url:
page += 1
params = {'start': start, 'rows': rows}
params.update(search_params)
logger.info(
f'Searching packages {url} PAGE:{page} start:{start}, rows:{rows} with params: {params}'
)
headers = self.get_request_headers()
try:
if method == 'POST': # depend on CKAN version
req = requests.post(url, data=params, headers=headers)
else:
req = requests.get(url, params=params, headers=headers)
except Exception as e:
error = 'ERROR Donwloading package list: {} [{}]'.format(
url, e)
raise ValueError(
'Failed to get package list at {}'.format(url))
content = req.content
if req.status_code >= 400:
error = ('ERROR searching CKAN package: {}'
'\n\t Status code: {}'
'\n\t Params: {}'
'\n\t content:{}'.format(url, req.status_code, params,
content))
logger.error(error)
raise Exception(error)
try:
json_content = json.loads(content) # check for encoding errors
except Exception as e:
error = 'ERROR parsing JSON data: {} [{}]'.format(content, e)
raise ValueError(error)
if not json_content['success']:
error = 'API response failed: {}'.format(
json_content.get('error', None))
raise ValueError(error)
result = json_content['result']
results = result['results']
real_results_count = len(results)
self.total_packages += real_results_count
logger.info(f'{real_results_count} results')
if real_results_count == 0:
url = None
else:
start += rows
self.package_list += results
logger.debug(f'datasets found: {results}')
yield (results)
