def process(self, item):
'''Generate a random dataset from a fake identifier'''
# Get or create a harvested dataset with this identifier.
# Harvest metadata are already filled on creation.
dataset = self.get_dataset(item.remote_id)
# Here you comes your implementation. You should :
# - fetch the remote dataset (if necessary)
# - validate the fetched payload
# - map its content to the dataset fields
# - store extra significant data in the `extra` attribute
# - map resources data
dataset.title = faker.sentence()
dataset.description = faker.text()
dataset.tags = list(set(faker.words(nb=faker.pyint())))
# Resources
for i in range(faker.pyint()):
dataset.resources.append(
Resource(title=faker.sentence(),
description=faker.text(),
url=faker.url(),
filetype='remote',
mime=faker.mime_type(category='text'),
format=faker.file_extension(category='text'),
filesize=faker.pyint()))
return dataset
def process(self, item):
kwargs = item.kwargs
item.remote_id = kwargs['remote_id']
dataset = self.get_dataset(item.remote_id)
dataset.title = kwargs['title']
dataset.tags = kwargs.get('tags', '').split(',')
dataset.private = True
dataset.frequency = kwargs.get('dataset_frequency', 'annual')
license = kwargs.get('dataset_license', 'cc-zero')
dataset.license = License.objects.get(id=license)
resources = kwargs['resources']
description = u"Ce jeu de données contient: <br>"
for resource in resources:
description += resource['title'] + "<br>"
dataset.description = description
# Force recreation of all resources
dataset.resources = []
for resource in resources:
new_resource = Resource(title=resource['title'],
url=resource['link'],
filetype=resource['filetype'],
format=resource['format'],
description=resource.get(
'description', ''))
dataset.resources.append(new_resource)
return dataset
def process(self, item):
dataset = self.get_dataset(item.remote_id)
# Here you comes your implementation. You should :
# - fetch the remote dataset (if necessary)
# - validate the fetched payload
# - map its content to the dataset fields
# - store extra significant data in the `extra` attribute
# - map resources data
kwargs = item.kwargs
dataset.title = kwargs['title']
dataset.tags = ["statec-harvesting"]
resources = kwargs['resources']
description = u"Ce jeu de données contient: <br>"
for resource in resources:
description += resource['title'] + "<br>"
description += "<br>---------------------------------------"
description += """<br> Automatically synched from
portail statistique (category %s)""" % dataset.title
dataset.description = description
# Force recreation of all resources
dataset.resources = []
for resource in resources:
url = resource['link']
url = url.replace('tableView', 'download')
params = {
'IF_DOWNLOADFORMAT': 'csv',
'IF_DOWNLOAD_ALL_ITEMS': 'yes'
}
url_parts = list(urlparse.urlparse(url))
query = dict(urlparse.parse_qsl(url_parts[4]))
query.update(params)
url_parts[4] = urlencode(query)
download_url = urlparse.urlunparse(url_parts)
new_resource = Resource(title=resource['title'],
url=download_url,
filetype='remote',
format='csv')
if len(
filter(
lambda d: d['title'] in [resource['title']] and d[
'url'] in [download_url],
dataset.resources)) == 0: # noqa
dataset.resources.append(new_resource)
else:
pass
return dataset
def process_resources(self, dataset, data, formats):
dataset_id = data["datasetid"]
ods_metadata = data["metas"]
description = self.description_from_fields(data['fields'])
for format in formats:
label, udata_format, mime = self.FORMATS[format]
resource = Resource(title='Export au format {0}'.format(label),
description=description,
filetype='remote',
url=self._get_download_url(dataset_id, format),
format=udata_format,
mime=mime)
resource.modified = ods_metadata["modified"]
dataset.resources.append(resource)
def process(self, item):
dataset = self.get_dataset(item.remote_id)
# Here you comes your implementation. You should :
# - fetch the remote dataset (if necessary)
# - validate the fetched payload
# - map its content to the dataset fields
# - store extra significant data in the `extra` attribute
# - map resources data
kwargs = item.kwargs
item = kwargs['item']
dataset.title = item['title']
dataset.license = License.guess('cc-by')
dataset.tags = ["snig.dgterritorio.gov.pt"]
dataset.description = item['description']
if item.get('date'):
dataset.created_at = item['date']
for keyword in item.get('keywords'):
dataset.tags.append(keyword)
# Force recreation of all resources
dataset.resources = []
for resource in item.get("resources"):
parsed = urlparse.urlparse(resource['url'])
try:
format = str(urlparse.parse_qs(parsed.query)['service'][0])
except KeyError:
format = resource['url'].split('.')[-1]
new_resource = Resource(title=item['title'],
url=resource['url'],
filetype='remote',
format=format)
dataset.resources.append(new_resource)
dataset.extras['harvest:name'] = self.source.name
return dataset
def process(self, item):
dataset = self.get_dataset(item.remote_id)
# Here you comes your implementation. You should :
# - fetch the remote dataset (if necessary)
# - validate the fetched payload
# - map its content to the dataset fields
# - store extra significant data in the `extra` attribute
# - map resources data
kwargs = item.kwargs
dataset.title = kwargs['title']
dataset.license = License.guess('cc-by')
dataset.tags = ["apambiente.pt"]
item = kwargs['item']
dataset.description = item['summary']
if kwargs['date']:
dataset.created_at = kwargs['date']
# Force recreation of all resources
dataset.resources = []
for resource in item['links']:
url = resource['href'].replace('\\', '').replace(' ', '%20')
type = resource['type']
if type == 'details':
dataset.description += "<br>"
dataset.description += "<br>Mais detalhes : <a href=\"%s\" target=\"_blank\">%s</a>" % (
url, dataset.title)
if type == 'open':
url_parts = list(urlparse.urlparse(url))
parts = url_parts[2].split('.')
format = parts[-1] if len(parts) > 1 else 'wms'
new_resource = Resource(title=dataset.title,
url=url,
filetype='remote',
format=format.lower())
dataset.resources.append(new_resource)
return dataset
def process(self, item):
dataset = self.get_dataset(item.remote_id)
# Here you comes your implementation. You should :
# - fetch the remote dataset (if necessary)
# - validate the fetched payload
# - map its content to the dataset fields
# - store extra significant data in the `extra` attribute
# - map resources data
kwargs = item.kwargs
dataset.title = kwargs['title']
dataset.license = License.guess('cc-by')
dataset.tags = ["apambiente.pt"]
item = kwargs['item']
dataset.description = item.get('description')
if kwargs['date']:
dataset.created_at = kwargs['date']
# Force recreation of all resources
dataset.resources = []
url = item.get('url')
if item.get('type') == "liveData":
type = "wms"
else:
type = url.split('.')[-1].lower()
if len(type) > 3:
type = "wms"
new_resource = Resource(title=dataset.title,
url=url,
filetype='remote',
format=type)
dataset.resources.append(new_resource)
return dataset
def process(self, item):
# dataset_id is found in item.remote_id
dataset_id = item.remote_id
# we want to avoid a same id from another
# source so we add 'instn' in front of it
dataset = self.get_dataset('instn-%s' % item.remote_id)
# We saved the theme name as a tag and the dataset name
# as the title
dataset.title = item.kwargs['dataset_name']
dataset.tags = [item.kwargs['theme_name']]
# We empty the existing resources (= the different "files" for this dataset)
# Here we will have only one probably
dataset.resources = []
# Getting the xls file url
dataset_request = requests.get(
'http://beta.ins.tn/fr/node/get/nojs/%s' % dataset_id)
dataset_soup = BeautifulSoup(dataset_request.content, 'html.parser')
dataset_xls = dataset_soup.find_all('li', {"class": "data"})
for xls in dataset_xls:
url = xls.find('a').attrs['href']
# We have the url, let's do a HEAD request
# to get the file size without downloading the whole file
size_request = requests.head(url)
file_size = size_request.headers['Content-length']
# We're good, let's add the file
# to the dataset resources
dataset.resources.append(
Resource(
title=dataset.title,
url=url,
filetype='remote',
mime=
'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet',
format='xlsx',
filesize=file_size))
# at the end, always return the dataset
return dataset
def get_resource(self, dataset, url):
resource = get_by(dataset.resources, 'url', url)
if not resource:
return True, Resource(url=url)
return False, resource
def process(self, item):
response = self.get_action('package_show', id=item.remote_id)
data = self.validate(response['result'], schema)
# Fix the remote_id: use real ID instead of not stable name
item.remote_id = data['id']
# Skip if no resource
if not len(data.get('resources', [])):
msg = 'Dataset {0} has no record'.format(item.remote_id)
raise HarvestSkipException(msg)
dataset = self.get_dataset(item.remote_id)
# Core attributes
if not dataset.slug:
dataset.slug = data['name']
dataset.title = data['title']
dataset.description = data['notes']
dataset.license = License.objects(id=data['license_id']).first()
# dataset.license = license or License.objects.get(id='notspecified')
dataset.tags = [t['name'] for t in data['tags'] if t['name']]
dataset.created_at = data['metadata_created']
dataset.last_modified = data['metadata_modified']
dataset.extras['ckan:name'] = data['name']
temporal_start, temporal_end = None, None
spatial_geom = None
for extra in data['extras']:
# GeoJSON representation (Polygon or Point)
if extra['key'] == 'spatial':
spatial_geom = json.loads(extra['value'])
# Textual representation of the extent / location
elif extra['key'] == 'spatial-text':
log.debug('spatial-text value not handled')
print 'spatial-text', extra['value']
# Linked Data URI representing the place name
elif extra['key'] == 'spatial-uri':
log.debug('spatial-uri value not handled')
print 'spatial-uri', extra['value']
# Update frequency
elif extra['key'] == 'frequency':
print 'frequency', extra['value']
# Temporal coverage start
elif extra['key'] == 'temporal_start':
print 'temporal_start', extra['value']
temporal_start = daterange_start(extra['value'])
continue
# Temporal coverage end
elif extra['key'] == 'temporal_end':
print 'temporal_end', extra['value']
temporal_end = daterange_end(extra['value'])
continue
# else:
# print extra['key'], extra['value']
dataset.extras[extra['key']] = extra['value']
if spatial_geom:
dataset.spatial = SpatialCoverage()
if spatial_geom['type'] == 'Polygon':
coordinates = [spatial_geom['coordinates']]
elif spatial_geom['type'] == 'MultiPolygon':
coordinates = spatial_geom['coordinates']
else:
HarvestException('Unsupported spatial geometry')
dataset.spatial.geom = {
'type': 'MultiPolygon',
'coordinates': coordinates
}
if temporal_start and temporal_end:
dataset.temporal_coverage = db.DateRange(
start=temporal_start,
end=temporal_end,
)
# Remote URL
if data.get('url'):
dataset.extras['remote_url'] = data['url']
# Resources
for res in data['resources']:
if res['resource_type'] not in ALLOWED_RESOURCE_TYPES:
continue
try:
resource = get_by(dataset.resources, 'id', UUID(res['id']))
except:
log.error('Unable to parse resource ID %s', res['id'])
continue
if not resource:
resource = Resource(id=res['id'])
dataset.resources.append(resource)
resource.title = res.get('name', '') or ''
resource.description = res.get('description')
resource.url = res['url']
resource.filetype = ('api' if res['resource_type'] == 'api'
else 'remote')
resource.format = res.get('format')
resource.mime = res.get('mimetype')
resource.hash = res.get('hash')
resource.created = res['created']
resource.modified = res['last_modified']
resource.published = resource.published or resource.created
return dataset
def process(self, item):
dataset = self.get_dataset(item.remote_id)
# Here you comes your implementation. You should :
# - fetch the remote dataset (if necessary)
# - validate the fetched payload
# - map its content to the dataset fields
# - store extra significant data in the `extra` attribute
# - map resources data
# check if this is a new dataset and give it a title
if dataset.title is None:
dataset.title = ''
pass
# Create the new list of tags and mmake sure the list has only unique tags
tags = []
for tag in dataset.tags:
tags.append(tag)
pass
tags.append("statec-harvesting")
tags = list(set(tags))
dataset.tags = tags
# return the gathered resources of the items
# or return the updated list of all the resources of the given dataset
resources = self.__update_resources(item, dataset)
if dataset.title == '':
dataset.title = item.kwargs['title']
# Rebuild the dataset description
description = u"This dataset includes the following resource(s): <br>"
for resource in resources:
description += resource.title + "<br>"
description += "<br>---------------------------------------"
description += """<br> Automatically synched from
portail statistique (category %s)""" % dataset.title
dataset.description = description
# Force recreation of all resources
dataset.resources = []
for resource in resources:
url = resource.url
download_url = url
# check if the resource format is csv and handle the link creation accordingly
if resource.format == 'csv':
url = url.replace('tableView', 'download')
params = {
'IF_DOWNLOADFORMAT': 'csv',
'IF_DOWNLOAD_ALL_ITEMS': 'yes'
}
url_parts = list(urlparse.urlparse(url))
query = dict(urlparse.parse_qsl(url_parts[4]))
query.update(params)
url_parts[4] = urlencode(query)
download_url = urlparse.urlunparse(url_parts)
pass
# The newly created resource
new_resource = Resource(
title=resource.title,
description=resource.title,
url=download_url,
filetype='remote',
format=resource.format
)
dataset.resources.append(new_resource)
return dataset
def remote_datasets(self):
response = self.get('package_list')
for name in response['result']:
details = self.get('package_show', {'id': name})['result']
dataset = self.get_harvested(Dataset, details['id'])
# Core attributes
dataset.slug = details['name']
dataset.title = details['title']
dataset.description = details.get('notes', 'No description')
dataset.license = License.objects(
id=details['license_id']).first() or License.objects.get(
id='notspecified')
dataset.tags = [tag['name'].lower() for tag in details['tags']]
dataset.frequency = self.map('frequency', details) or 'unknown'
dataset.created_at = parse(details['metadata_created'])
dataset.last_modified = parse(details['metadata_modified'])
if any_field(details, 'territorial_coverage',
'territorial_coverage_granularity'):
coverage = TerritorialCoverage(
codes=[
code.strip() for code in details.get(
'territorial_coverage', '').split(',')
if code.strip()
],
granularity=self.map('territorial_coverage_granularity',
details),
)
dataset.extras['territorial_coverage'] = coverage
try:
dataset.spatial = territorial_to_spatial(dataset)
except Exception as e:
print 'Error while processing spatial coverage for {0}:'.format(
dataset.title), e
if all_field(details, 'temporal_coverage_from',
'temporal_coverage_to'):
try:
dataset.temporal_coverage = db.DateRange(
start=daterange_start(
details.get('temporal_coverage_from')),
end=daterange_end(details.get('temporal_coverage_to')),
)
except:
log.error(
'Unable to parse temporal coverage for dataset %s',
details['id'])
# Organization
if details.get('organization'):
dataset.organization = self.get_harvested(
Organization, details['organization']['id'], False)
else:
# Need to fetch user from roles
roles = self.get('roles_show',
{'domain_object': name})['result']['roles']
for role in roles:
if role['role'] == 'admin' and role['context'] == 'Package':
dataset.owner = self.get_harvested(
User, role['user_id'])
break
# Supplier
if details.get('supplier_id'):
dataset.supplier = self.get_harvested(Organization,
details['supplier_id'],
False)
# Remote URL
if details.get('url'):
dataset.extras['remote_url'] = details['url']
# Extras
if 'extras' in details:
extra_mapping = self.harvester.mapping.get('from_extras', {})
for extra in details['extras']:
if extra['key'] in self.harvester.mapping:
value = self.harvester.mapping[extra['key']].get(
extra['value'])
else:
value = extra['value']
if extra['key'] in extra_mapping:
setattr(dataset, extra_mapping[extra['key']], value)
else:
dataset.extras[extra['key']] = value
# Resources
for res in details['resources']:
try:
resource = get_by(dataset.resources, 'id', UUID(res['id']))
except:
log.error('Unable to parse resource %s', res['id'])
continue
if not resource:
resource = Resource(id=res['id'])
dataset.resources.append(resource)
resource.title = res.get('name', '') or ''
resource.url = res['url']
resource.description = res.get('description')
resource.format = res.get('format')
resource.hash = res.get('hash')
resource.created = parse(res['created'])
resource.modified = parse(res['revision_timestamp'])
resource.published = resource.published or resource.created
yield dataset
if dataset.id:
followers = self.get('dataset_follower_list',
{'id': name})['result']
for follower in followers:
user = self.get_harvested(User, follower['id'], False)
if user:
follow, created = FollowDataset.objects.get_or_create(
follower=user, following=dataset)
def process(self, item):
'''Return the INE datasets'''
dataset = self.get_dataset(item.remote_id)
# get remote data for dataset
req = requests.get("https://www.ine.pt/ine/xml_indic.jsp",
params={
'varcd': item.remote_id,
'lang': 'PT',
'opc': '1'
},
headers={'charset': 'utf8'})
returnedData = req.content
print 'Get metadata for %s' % (item.remote_id)
keywordSet = set()
dataset.license = License.guess('cc-by')
dataset.resources = []
doc = minidom.parseString(returnedData)
properties = doc.getElementsByTagName('indicator')
# go through the API dataset information
for propNode in properties:
for childNode in propNode.childNodes:
# print childNode
fc = childNode.firstChild
if fc:
if childNode.nodeName == 'keywords':
for obj in childNode.childNodes:
# INE needs to create a proper xml file...
valueData = obj.nodeValue
# need to ignore the ',' nodes
if obj.nodeValue != ',':
# need to ignore the last "," usually after the INE value
if valueData[-1:] == ',':
valueData = valueData[:-1]
# this removes redundant keywords that sometimes show with different cases (lower and upper)
keywordSet.add(valueData.lower())
dataset.tags = list(keywordSet)
dataset.tags.append('ine.pt')
dataset.frequency = 'unknown'
elif childNode.nodeName == 'title':
dataset.title = fc.nodeValue
elif childNode.nodeName == 'description':
dataset.description = fc.nodeValue
elif childNode.nodeName == 'html':
for obj in childNode.childNodes:
if obj.nodeName == 'bdd_url':
dataset.description += "\n " + obj.firstChild.nodeValue
elif childNode.nodeName == 'json':
for obj in childNode.childNodes:
if obj.nodeName == 'json_dataset':
dataset.resources.append(
Resource(
title='Dataset json url',
description='Dataset em formato json',
url=obj.firstChild.nodeValue,
filetype='remote',
format='json'))
elif obj.nodeName == 'json_metainfo':
dataset.resources.append(
Resource(
title='Json metainfo url',
description='Metainfo em formato json',
url=obj.firstChild.nodeValue,
filetype='remote',
format='json'))
return dataset
def process(self, item):
response = self.get(item.remote_id)
encoding = chardet.detect(response.content)['encoding']
xml = self.parse_xml(response.content.decode(encoding))
metadata = xml['metadata']
# Resolve and remote id from metadata
item.remote_id = metadata['id']
dataset = self.get_dataset(metadata['id'])
dataset.title = metadata['title']
dataset.frequency = FREQUENCIES.get(metadata['frequency'], 'unknown')
dataset.description = metadata['notes']
dataset.private = metadata['private']
dataset.tags = sorted(set(metadata['tags']))
if metadata.get('license_id'):
dataset.license = License.objects.get(id=metadata['license_id'])
if (metadata.get('temporal_coverage_from')
and metadata.get('temporal_coverage_to')):
dataset.temporal_coverage = db.DateRange(
start=metadata['temporal_coverage_from'],
end=metadata['temporal_coverage_to'])
if (metadata.get('territorial_coverage_code')
or metadata.get('territorial_coverage_granularity')):
dataset.spatial = SpatialCoverage()
if metadata.get('territorial_coverage_granularity'):
dataset.spatial.granularity = GRANULARITIES.get(
metadata['territorial_coverage_granularity'])
if metadata.get('territorial_coverage_code'):
dataset.spatial.zones = [
ZONES[metadata['territorial_coverage_code']]
]
dataset.resources = []
cle = get_by(metadata['resources'], 'format', 'cle')
for row in metadata['resources']:
if row['format'] == 'cle':
continue
else:
resource = Resource(title=row['name'],
description=(row['description'] + '\n\n' +
SSL_COMMENT).strip(),
filetype='remote',
url=row['url'],
format=row['format'])
if resource.format == 'csv' and cle:
resource.checksum = Checksum(type='sha256',
value=self.get(
cle['url']).text)
if row.get('last_modified'):
resource.modified = row['last_modified']
dataset.resources.append(resource)
if metadata.get('author'):
dataset.extras['author'] = metadata['author']
if metadata.get('author_email'):
dataset.extras['author_email'] = metadata['author_email']
if metadata.get('maintainer'):
dataset.extras['maintainer'] = metadata['maintainer']
if metadata.get('maintainer_email'):
dataset.extras['maintainer_email'] = metadata['maintainer_email']
for extra in metadata['extras']:
dataset.extras[extra['key']] = extra['value']
return dataset
def on_form_valid(self, form):
resource = Resource()
form.populate_obj(resource)
self.dataset.add_community_resource(resource)
return redirect(url_for('datasets.show', dataset=self.dataset))
def process(self, item):
response = self.get_action('package_show', id=item.remote_id)
data = self.validate(response['result'], self.schema)
if type(data) == list:
data = data[0]
# Fix the remote_id: use real ID instead of not stable name
item.remote_id = data['id']
# Skip if no resource
if not len(data.get('resources', [])):
msg = 'Dataset {0} has no record'.format(item.remote_id)
raise HarvestSkipException(msg)
dataset = self.get_dataset(item.remote_id)
# Core attributes
if not dataset.slug:
dataset.slug = data['name']
dataset.title = data['title']
dataset.description = parse_html(data['notes'])
# Detect Org
organization_acronym = data['organization']['name']
orgObj = Organization.objects(acronym=organization_acronym).first()
if orgObj:
#print 'Found %s' % orgObj.acronym
dataset.organization = orgObj
else:
orgObj = Organization()
orgObj.acronym = organization_acronym
orgObj.name = data['organization']['title']
orgObj.description = data['organization']['description']
orgObj.save()
#print 'Created %s' % orgObj.acronym
dataset.organization = orgObj
# Detect license
default_license = self.harvest_config.get('license', License.default())
dataset.license = License.guess(data['license_id'],
data['license_title'],
default=default_license)
dataset.tags = [t['name'] for t in data['tags'] if t['name']]
dataset.tags.append(urlparse(self.source.url).hostname)
dataset.created_at = data['metadata_created']
dataset.last_modified = data['metadata_modified']
dataset.frequency = 'unknown'
dataset.extras['ckan:name'] = data['name']
temporal_start, temporal_end = None, None
spatial_geom = None
for extra in data['extras']:
# GeoJSON representation (Polygon or Point)
if extra['key'] == 'spatial':
spatial_geom = json.loads(extra['value'])
# Textual representation of the extent / location
elif extra['key'] == 'spatial-text':
log.debug('spatial-text value not handled')
# Linked Data URI representing the place name
elif extra['key'] == 'spatial-uri':
log.debug('spatial-uri value not handled')
# Update frequency
elif extra['key'] == 'frequency':
print 'frequency', extra['value']
# Temporal coverage start
elif extra['key'] == 'temporal_start':
temporal_start = daterange_start(extra['value'])
continue
# Temporal coverage end
elif extra['key'] == 'temporal_end':
temporal_end = daterange_end(extra['value'])
continue
dataset.extras[extra['key']] = extra['value']
# We don't want spatial to be added on harvester
if self.harvest_config.get('geozones', False):
dataset.spatial = SpatialCoverage()
dataset.spatial.zones = []
for zone in self.harvest_config.get('geozones'):
geo_zone = GeoZone.objects.get(id=zone)
dataset.spatial.zones.append(geo_zone)
#
# if spatial_geom:
# dataset.spatial = SpatialCoverage()
# if spatial_geom['type'] == 'Polygon':
# coordinates = [spatial_geom['coordinates']]
# elif spatial_geom['type'] == 'MultiPolygon':
# coordinates = spatial_geom['coordinates']
# else:
# HarvestException('Unsupported spatial geometry')
# dataset.spatial.geom = {
# 'type': 'MultiPolygon',
# 'coordinates': coordinates
# }
if temporal_start and temporal_end:
dataset.temporal_coverage = db.DateRange(
start=temporal_start,
end=temporal_end,
)
# Remote URL
if data.get('url'):
try:
url = uris.validate(data['url'])
except uris.ValidationError:
dataset.extras['remote_url'] = self.dataset_url(data['name'])
dataset.extras['ckan:source'] = data['url']
else:
dataset.extras['remote_url'] = url
dataset.extras['harvest:name'] = self.source.name
current_resources = [
str(resource.id) for resource in dataset.resources
]
fetched_resources = []
# Resources
for res in data['resources']:
if res['resource_type'] not in ALLOWED_RESOURCE_TYPES:
continue
#Ignore invalid Resources
try:
url = uris.validate(res['url'])
except uris.ValidationError:
continue
try:
resource = get_by(dataset.resources, 'id', UUID(res['id']))
except Exception:
log.error('Unable to parse resource ID %s', res['id'])
continue
fetched_resources.append(str(res['id']))
if not resource:
resource = Resource(id=res['id'])
dataset.resources.append(resource)
resource.title = res.get('name', '') or ''
resource.description = parse_html(res.get('description'))
resource.url = res['url']
resource.filetype = 'remote'
resource.format = res.get('format')
resource.mime = res.get('mimetype')
resource.hash = res.get('hash')
resource.created = res['created']
resource.modified = res['last_modified']
resource.published = resource.published or resource.created
# Clean up old resources removed from source
for resource_id in current_resources:
if resource_id not in fetched_resources:
try:
resource = get_by(dataset.resources, 'id',
UUID(resource_id))
except Exception:
log.error('Unable to parse resource ID %s', resource_id)
continue
else:
if resource and not self.dryrun:
dataset.resources.remove(resource)
return dataset
def process(self, item):
'''Return the DadosGov datasets with the corresponding original and xml file'''
global REPORT_FILE_PATH, DADOSGOVPATH, DOWNLOADFILEPATH, DADOSGOVURL
reload(sys)
sys.setdefaultencoding('utf8')
# Get or create a harvested dataset with this identifier.
dataset = self.get_dataset(item.remote_id)
# get the organization object, no check necessary, it should always exist
orgObj = Organization.objects(id=item.kwargs['orgId']).first()
print '------------------------------------'
print 'Processing %s (%s)' % (dataset.title, item.remote_id)
# print item.kwargs
# print '--'
# set additional vars
dataset.tags = ['migrado']
dataset.extras = {}
dataset.organization = orgObj.id
dataset.license = License.guess('cc-by')
dataset.resources = []
# *********************************************
# go through the DB dataset information
dataset.created_at = item.kwargs['createdOn']
dataset.extras['links'] = item.kwargs['serviceUrl']
# ********************************************************
# ********************************************************
req = requests.get(
"http://%s/v1/%s/TableMetadata" %
(DADOSGOVURL, item.kwargs['orgAcronym']),
params={'$filter': "partitionkey eq '%s'" % item.remote_id},
headers={'charset': 'utf8'})
xmlRootData = req.content
propertiesDoc = minidom.parseString(xmlRootData)
propertiesStuff = propertiesDoc.getElementsByTagName('content')
propEl = propertiesDoc.getElementsByTagNameNS('*', 'properties')
if propEl:
propertiesElements = propEl[0].childNodes
# go through the API dataset information
for propEl in propertiesElements:
if propEl.nodeType == Node.ELEMENT_NODE:
fc = propEl.firstChild
if fc:
if propEl.nodeName == 'd:category':
dataset.tags.append(fc.nodeValue)
elif propEl.nodeName == 'd:keywords':
dataset.tags.extend([
currentTag.strip()
for currentTag in fc.nodeValue.split(',')
])
# elif propEl.nodeName == 'd:PartitionKey':
# dataset.slug = fc.nodeValue
elif propEl.nodeName == 'd:nameexternal':
dataset.title = fc.nodeValue
elif propEl.nodeName == 'd:description':
dataset.description = fc.nodeValue
elif propEl.nodeName == 'd:contact':
dataset.extras['contact'] = fc.nodeValue
elif propEl.nodeName == 'd:links' and fc.nodeValue:
dataset.extras['links'] = '%s, %s' % (
dataset.extras['links'], fc.nodeValue)
# ********************************************************
env = current_app.config.get('MIGRATION_URL')
if env:
fixedUrl = env
else:
fixedUrl = url_for('site.home', _external=True)
fixedUrl = '%s/s/%s' % (fixedUrl[:fixedUrl.rfind('/', 0, -1)],
DADOSGOVPATH)
# empty previous dataset resources
dataset.resources = []
# separate filename from extension
filename = os.path.splitext(item.kwargs['filePath'])
# ********************************************************
# get xml by api and set the dataset resource field:
# filenameXml = '%s.xml' % (filename[0])
filenameXml = '%s.xml' % (item.remote_id)
u = urllib2.urlopen(
"http://%s/v1/%s/%s" %
(DADOSGOVURL, item.kwargs['orgAcronym'], item.remote_id))
# create/open the local file to be written
with open('%s/%s' % (DOWNLOADFILEPATH, filenameXml), 'wb') as f:
# write file data
f.write(u.read())
# get file size info
meta = u.info()
fileSize = int(meta.getheaders("Content-Length")[0])
fullPath = '%s/%s' % (fixedUrl, filenameXml)
print fullPath
# set the resource data for the dataset
dataset.resources.append(
Resource(title=dataset.title,
description='Dados em formato xml',
url=fullPath,
mime='text/xml ',
format='xml',
filesize=fileSize,
created_at=item.kwargs['createdOn']))
# ********************************************************
# ********************************************************
# get json by api and set the dataset resource field:
filenameJson = '%s.json' % (item.remote_id)
u = urllib2.urlopen(
"http://%s/v1/%s/%s?format=json" %
(DADOSGOVURL, item.kwargs['orgAcronym'], item.remote_id))
# create/open the local file to be written
with open('%s/%s' % (DOWNLOADFILEPATH, filenameJson), 'wb') as f:
# write file data
f.write(u.read())
# get file size info
meta = u.info()
fileSize = int(meta.getheaders("Content-Length")[0])
fullPath = '%s/%s' % (fixedUrl, filenameJson)
print fullPath
# set the resource data for the dataset
dataset.resources.append(
Resource(title=dataset.title,
description='Dados em formato json',
url=fullPath,
mime='application/json ',
format='json',
filesize=fileSize,
created_at=item.kwargs['createdOn']))
# ********************************************************
# ********************************************************
# get original files using static path and ftp and set the dataset resource field
if item.kwargs['filePath']:
try:
# https://dadosgovstorage.blob.core.windows.net/datasetsfiles/Acesso%20a%20Consultas%20M%C3%A9dicas%20pela%20Popula%C3%A7%C3%A3o%20Inscrita_636046701023924396.xlsx
print '-- ** filePath ** --> %s' % item.kwargs['filePath']
try:
urlSafe = urllib.quote(item.kwargs['filePath'])
print "https://dadosgovstorage.blob.core.windows.net/datasetsfiles/%s" % (
urlSafe)
u = urllib2.urlopen(
"https://dadosgovstorage.blob.core.windows.net/datasetsfiles/%s"
% (urlSafe))
# create/open the local file to be written
with open(
'%s/%s%s' %
(DOWNLOADFILEPATH, item.remote_id, filename[1]),
'wb') as f:
# write file data
f.write(u.read())
# get file size info
meta = u.info()
fileSize = int(
meta.getheaders("Content-Length")[0])
fullPath = '%s/%s%s' % (fixedUrl, item.remote_id,
filename[1])
print fullPath
# set the resource data for the dataset
dataset.resources.append(
Resource(title=dataset.title,
description='Ficheiro original (%s)' %
(item.kwargs['filePath']),
url=fullPath,
mime='application/vnd.ms-excel',
format=filename[1][1:],
filesize=fileSize,
created_at=item.kwargs['createdOn']))
except KeyError:
print '************ Error ************'
print traceback.format_exc()
print '*******************************'
# file not found exception
except IOError as ex:
print 'Original file not found:'
print ex
# ********************************************************
print '--'
print 'Returning %s' % dataset.title
print '------------------------------------'
with open(REPORT_FILE_PATH, 'a') as csvResFile:
writer = csv.writer(csvResFile,
delimiter=chr(9),
quotechar=chr(34),
quoting=csv.QUOTE_MINIMAL)
writer.writerow([
item.remote_id, dataset.title, orgObj.name,
json.dumps(dataset.tags, ensure_ascii=False),
item.kwargs['filePath'], filenameXml, '', '[]'
])
# update the number of datasets associated with this organization
orgObj.metrics['datasets'] += 1
orgObj.save()
return dataset
print 'No data returned from the API for the dataset %s' % (
item.remote_id)
with open(REPORT_FILE_PATH, 'a') as csvResFile:
writer = csv.writer(csvResFile,
delimiter=chr(9),
quotechar=chr(34),
quoting=csv.QUOTE_MINIMAL)
writer.writerow([
item.remote_id, '', '', '', item.kwargs['filePath'], '', '',
'[]'
])
return None
def process(self, item):
response = self.get_action('package_show', id=item.remote_id)
data = self.validate(response['result'], self.schema)
if type(data) == list:
data = data[0]
# Fix the remote_id: use real ID instead of not stable name
item.remote_id = data['id']
# Skip if no resource
if not len(data.get('resources', [])):
msg = 'Dataset {0} has no record'.format(item.remote_id)
raise HarvestSkipException(msg)
dataset = self.get_dataset(item.remote_id)
# Core attributes
if not dataset.slug:
dataset.slug = data['name']
dataset.title = data['title']
dataset.description = parse_html(data['notes'])
# Detect license
default_license = dataset.license or License.default()
dataset.license = License.guess(data['license_id'],
data['license_title'],
default=default_license)
dataset.tags = [t['name'] for t in data['tags'] if t['name']]
dataset.created_at = data['metadata_created']
dataset.last_modified = data['metadata_modified']
dataset.extras['ckan:name'] = data['name']
temporal_start, temporal_end = None, None
spatial_geom, spatial_zone = None, None
for extra in data['extras']:
key = extra['key']
value = extra['value']
if value is None or (isinstance(value, str) and not value.strip()):
# Skip empty extras
continue
elif key == 'spatial':
# GeoJSON representation (Polygon or Point)
spatial_geom = json.loads(value)
elif key == 'spatial-text':
# Textual representation of the extent / location
qs = GeoZone.objects(db.Q(name=value) | db.Q(slug=value))
qs = qs.valid_at(datetime.now())
if qs.count() == 1:
spatial_zone = qs.first()
else:
dataset.extras['ckan:spatial-text'] = value
log.debug('spatial-text value not handled: %s', value)
elif key == 'spatial-uri':
# Linked Data URI representing the place name
dataset.extras['ckan:spatial-uri'] = value
log.debug('spatial-uri value not handled: %s', value)
elif key == 'frequency':
# Update frequency
freq = frequency_from_rdf(value)
if freq:
dataset.frequency = freq
elif value in UPDATE_FREQUENCIES:
dataset.frequency = value
else:
dataset.extras['ckan:frequency'] = value
log.debug('frequency value not handled: %s', value)
# Temporal coverage start
elif key == 'temporal_start':
temporal_start = daterange_start(value)
# Temporal coverage end
elif key == 'temporal_end':
temporal_end = daterange_end(value)
else:
dataset.extras[extra['key']] = value
if spatial_geom or spatial_zone:
dataset.spatial = SpatialCoverage()
if spatial_zone:
dataset.spatial.zones = [spatial_zone]
if spatial_geom:
if spatial_geom['type'] == 'Polygon':
coordinates = [spatial_geom['coordinates']]
elif spatial_geom['type'] == 'MultiPolygon':
coordinates = spatial_geom['coordinates']
else:
raise HarvestException('Unsupported spatial geometry')
dataset.spatial.geom = {
'type': 'MultiPolygon',
'coordinates': coordinates
}
if temporal_start and temporal_end:
dataset.temporal_coverage = db.DateRange(
start=temporal_start,
end=temporal_end,
)
# Remote URL
dataset.extras['remote_url'] = self.dataset_url(data['name'])
if data.get('url'):
try:
url = uris.validate(data['url'])
except uris.ValidationError:
dataset.extras['ckan:source'] = data['url']
else:
# use declared `url` as `remote_url` if any
dataset.extras['remote_url'] = url
# Resources
for res in data['resources']:
if res['resource_type'] not in ALLOWED_RESOURCE_TYPES:
continue
try:
resource = get_by(dataset.resources, 'id', UUID(res['id']))
except Exception:
log.error('Unable to parse resource ID %s', res['id'])
continue
if not resource:
resource = Resource(id=res['id'])
dataset.resources.append(resource)
resource.title = res.get('name', '') or ''
resource.description = parse_html(res.get('description'))
resource.url = res['url']
resource.filetype = 'remote'
resource.format = res.get('format')
resource.mime = res.get('mimetype')
resource.hash = res.get('hash')
resource.created = res['created']
resource.modified = res['last_modified']
resource.published = resource.published or resource.created
return dataset