def validator(self):
# Validates that a URL is a good data.json file.
if request.method == "POST" and "url" in request.POST and request.POST["url"].strip() != "":
c.source_url = request.POST["url"]
c.errors = []
import urllib
import json
from datajsonvalidator import do_validation
body = None
try:
body = json.load(urllib.urlopen(c.source_url))
except IOError as e:
c.errors.append(("Error Loading File", ["The address could not be loaded: " + unicode(e)]))
except ValueError as e:
c.errors.append(("Invalid JSON", ["The file does not meet basic JSON syntax requirements: " + unicode(
e) + ". Try using JSONLint.com."]))
except Exception as e:
c.errors.append((
"Internal Error",
["Something bad happened while trying to load and parse the file: " + unicode(e)]))
if body:
try:
do_validation(body, c.errors)
except Exception as e:
c.errors.append(("Internal Error", ["Something bad happened: " + unicode(e)]))
if len(c.errors) == 0:
c.errors.append(("No Errors", ["Great job!"]))
return render('datajsonvalidator.html')
def validate(pkg, dataset_dict):
import sys, os
global currentPackageOrg
try:
# When saved from UI DataQuality value is stored as "on" instead of True.
# Check if value is "on" and replace it with True.
dataset_dict = OrderedDict(dataset_dict)
if dataset_dict.get('dataQuality') == "on" \
or dataset_dict.get('dataQuality') == "true" \
or dataset_dict.get('dataQuality') == "True":
dataset_dict['dataQuality'] = True
elif dataset_dict.get('dataQuality') == "false" \
or dataset_dict.get('dataQuality') == "False":
dataset_dict['dataQuality'] = False
errors = []
try:
from datajsonvalidator import do_validation
do_validation([dict(dataset_dict)], errors,
Package2Pod.seen_identifiers)
except Exception as e:
errors.append(("Internal Error",
["Something bad happened: " + unicode(e)]))
if len(errors) > 0:
for error in errors:
log.warn(error)
try:
currentPackageOrg
except NameError:
currentPackageOrg = 'unknown'
errors_dict = OrderedDict([
('id', pkg.get('id')),
('name', Package2Pod.filter(pkg.get('name'))),
('title', Package2Pod.filter(pkg.get('title'))),
('organization', Package2Pod.filter(currentPackageOrg)),
('errors', errors),
])
return errors_dict
return dataset_dict
except Exception as e:
exc_type, exc_obj, exc_tb = sys.exc_info()
filename = os.path.split(exc_tb.tb_frame.f_code.co_filename)[1]
log.error("%s : %s : %s", exc_type, filename, exc_tb.tb_lineno)
raise e
def validate(pkg, dataset_dict):
import sys, os
global currentPackageOrg
try:
# When saved from UI DataQuality value is stored as "on" instead of True.
# Check if value is "on" and replace it with True.
dataset_dict = OrderedDict(dataset_dict)
if dataset_dict.get('dataQuality') == "on" \
or dataset_dict.get('dataQuality') == "true" \
or dataset_dict.get('dataQuality') == "True":
dataset_dict['dataQuality'] = True
elif dataset_dict.get('dataQuality') == "false" \
or dataset_dict.get('dataQuality') == "False":
dataset_dict['dataQuality'] = False
errors = []
try:
from datajsonvalidator import do_validation
do_validation([dict(dataset_dict)], errors, Package2Pod.seen_identifiers)
except Exception as e:
errors.append(("Internal Error", ["Something bad happened: " + unicode(e)]))
if len(errors) > 0:
for error in errors:
log.warn(error)
try:
currentPackageOrg
except NameError:
currentPackageOrg = 'unknown'
errors_dict = OrderedDict([
('id', pkg.get('id')),
('name', Package2Pod.filter(pkg.get('name'))),
('title', Package2Pod.filter(pkg.get('title'))),
('organization', Package2Pod.filter(currentPackageOrg)),
('errors', errors),
])
return errors_dict
return dataset_dict
except Exception as e:
exc_type, exc_obj, exc_tb = sys.exc_info()
filename = os.path.split(exc_tb.tb_frame.f_code.co_filename)[1]
log.error("%s : %s : %s", exc_type, filename, exc_tb.tb_lineno)
raise e
def make_datajson_export_entry(package, seen_identifiers):
global currentPackageOrg
currentPackageOrg = None
# extras is a list of dicts [{},{}, {}]. For each dict, extract the key, value entries into a new dict
extras = dict([(x['key'], x['value']) for x in package['extras']])
parent_dataset_id = extras.get('parent_dataset')
if parent_dataset_id:
parent = model.Package.get(parent_dataset_id)
parent_uid = parent.extras.col.target['unique_id'].value
if parent_uid:
parent_dataset_id = parent_uid
# if resource format is CSV then convert it to text/csv
# Resource format has to be in 'csv' format for automatic datastore push.
for r in package["resources"]:
if r["format"].lower() == "csv":
r["format"] = "text/csv"
if r["format"].lower() == "json":
r["format"] = "application/json"
if r["format"].lower() == "pdf":
r["format"] = "application/pdf"
try:
retlist = [
("@type", "dcat:Dataset"), # optional
("title", JsonExportBuilder.strip_if_string(package["title"])), # required
# ("accessLevel", 'public'), # required
("accessLevel", JsonExportBuilder.strip_if_string(extras.get('public_access_level'))), # required
# ("accrualPeriodicity", "R/P1Y"), # optional
# ('accrualPeriodicity', 'accrual_periodicity'),
('accrualPeriodicity', JsonExportBuilder.get_accrual_periodicity(extras.get('accrual_periodicity'))),
# optional
("conformsTo", JsonExportBuilder.strip_if_string(extras.get('conforms_to'))), # optional
# ('contactPoint', OrderedDict([
# ("@type", "vcard:Contact"),
# ("fn", "Jane Doe"),
# ("hasEmail", "mailto:[email protected]")
# ])), # required
('contactPoint', JsonExportBuilder.get_contact_point(extras)), # required
("dataQuality", JsonExportBuilder.strip_if_string(extras.get('data_quality'))),
# required-if-applicable
("describedBy", JsonExportBuilder.strip_if_string(extras.get('data_dictionary'))), # optional
("describedByType", JsonExportBuilder.strip_if_string(extras.get('data_dictionary_type'))), # optional
("description", JsonExportBuilder.strip_if_string(package["notes"])), # required
# ("description", 'asdfasdf'), # required
("identifier", JsonExportBuilder.strip_if_string(extras.get('unique_id'))), # required
# ("identifier", 'asdfasdfasdf'), # required
("isPartOf", parent_dataset_id), # optional
("issued", JsonExportBuilder.strip_if_string(extras.get('release_date'))), # optional
# ("keyword", ['a', 'b']), # required
("keyword", [t["display_name"] for t in package["tags"]]), # required
("landingPage", JsonExportBuilder.strip_if_string(extras.get('homepage_url'))), # optional
("license", JsonExportBuilder.strip_if_string(extras.get("license_new"))), # required-if-applicable
("modified",
JsonExportBuilder.strip_if_string(extras.get("modified", package.get("metadata_modified")))),
# required
("primaryITInvestmentUII", JsonExportBuilder.strip_if_string(extras.get('primary_it_investment_uii'))),
# optional
# ('publisher', OrderedDict([
# ("@type", "org:Organization"),
# ("name", "Widget Services")
# ])), # required
# ("publisher", get_publisher_tree(extras)), # required
("publisher", JsonExportBuilder.get_publisher_tree_wrong_order(extras)), # required
("rights", JsonExportBuilder.strip_if_string(extras.get('access_level_comment'))), # required
("spatial", JsonExportBuilder.strip_if_string(package.get("spatial"))), # required-if-applicable
('systemOfRecords', JsonExportBuilder.strip_if_string(extras.get('system_of_records'))), # optional
("temporal", JsonExportBuilder.strip_if_string(extras.get('temporal'))), # required-if-applicable
("distribution", JsonExportBuilder.generate_distribution(package)), # required-if-applicable
# ("distribution",
# #TODO distribution should hide any key/value pairs where value is "" or None (e.g. format)
# [
# OrderedDict([
# ("downloadURL", r["url"]),
# ("mediaType", r["formatReadable"]),
# ])
# for r in package["resources"]
# ])
]
for pair in [
#('bureauCode', 'bureau_code'), # required
('language', 'language'), # optional
#('programCode', 'program_code'), # required
('references', 'related_documents'), # optional
('theme', 'category'), # optional
]:
JsonExportBuilder.split_multiple_entries(retlist, extras, pair)
except KeyError as e:
log.warn("Missing Required Field for package with id=[%s], title=['%s'], organization=['%s']: '%s'" % (
package.get('id'), package.get('title'), currentPackageOrg, e))
errors = ['Missing Required Field', ["%s" % e]]
errors_dict = OrderedDict([
('id', package.get('id')),
('name', package.get('name')),
('title', package.get('title')),
('organization', currentPackageOrg),
('errors', errors),
])
return errors_dict
# Remove entries where value is None, "", or empty list []
striped_retlist = [(x, y) for x, y in retlist if y is not None and y != "" and y != []]
# When saved from UI DataQuality value is stored as "on" instead of True.
# Check if value is "on" and replace it with True.
striped_retlist_dict = OrderedDict(striped_retlist)
if striped_retlist_dict.get('dataQuality') == "on" \
or striped_retlist_dict.get('dataQuality') == "true" \
or striped_retlist_dict.get('dataQuality') == "True":
striped_retlist_dict['dataQuality'] = True
elif striped_retlist_dict.get('dataQuality') == "false" \
or striped_retlist_dict.get('dataQuality') == "False":
striped_retlist_dict['dataQuality'] = False
errors = []
try:
do_validation([dict(striped_retlist_dict)], errors, seen_identifiers)
except Exception as e:
errors.append(("Internal Error", ["Something bad happened: " + unicode(e)]))
if len(errors) > 0:
for error in errors:
log.warn(error)
errors_dict = OrderedDict([
('id', package.get('id')),
('name', package.get('name')),
('title', package.get('title')),
('organization', currentPackageOrg),
('errors', errors),
])
return errors_dict
return striped_retlist_dict
def make_datajson_export_entry(package):
global currentPackageOrg
currentPackageOrg = None
# extras is a list of dicts [{},{}, {}]. For each dict, extract the key, value entries into a new dict
extras = dict([(x['key'], x['value']) for x in package['extras']])
parent_dataset_id = extras.get('parent_dataset')
if parent_dataset_id:
parent = model.Package.get(parent_dataset_id)
parent_uid = parent.extras.col.target['unique_id'].value
if parent_uid:
parent_dataset_id = parent_uid
# if resource format is CSV then convert it to text/csv
# Resource format has to be in 'csv' format for automatic datastore push.
for r in package["resources"]:
if r["format"].lower() == "csv":
r["format"] = "text/csv"
if r["format"].lower() == "json":
r["format"] = "application/json"
if r["format"].lower() == "pdf":
r["format"] = "application/pdf"
try:
retlist = [
("@type", "dcat:Dataset"), # optional
("title", JsonExportBuilder.strip_if_string(package["title"])), # required
# ("accessLevel", 'public'), # required
("accessLevel", JsonExportBuilder.strip_if_string(extras.get('public_access_level'))), # required
# ("accrualPeriodicity", "R/P1Y"), # optional
# ('accrualPeriodicity', 'accrual_periodicity'),
('accrualPeriodicity', JsonExportBuilder.get_accrual_periodicity(extras.get('accrual_periodicity'))),
# optional
("conformsTo", JsonExportBuilder.strip_if_string(extras.get('conforms_to'))), # optional
# ('contactPoint', OrderedDict([
# ("@type", "vcard:Contact"),
# ("fn", "Jane Doe"),
# ("hasEmail", "mailto:[email protected]")
# ])), # required
('contactPoint', JsonExportBuilder.get_contact_point(extras)), # required
("dataQuality", JsonExportBuilder.strip_if_string(extras.get('data_quality'))),
# required-if-applicable
("describedBy", JsonExportBuilder.strip_if_string(extras.get('data_dictionary'))), # optional
("describedByType", JsonExportBuilder.strip_if_string(extras.get('data_dictionary_type'))), # optional
("description", JsonExportBuilder.strip_if_string(package["notes"])), # required
# ("description", 'asdfasdf'), # required
("identifier", JsonExportBuilder.strip_if_string(extras.get('unique_id'))), # required
# ("identifier", 'asdfasdfasdf'), # required
("isPartOf", parent_dataset_id), # optional
("issued", JsonExportBuilder.strip_if_string(extras.get('release_date'))), # optional
# ("keyword", ['a', 'b']), # required
("keyword", [t["display_name"] for t in package["tags"]]), # required
("landingPage", JsonExportBuilder.strip_if_string(extras.get('homepage_url'))), # optional
("license", JsonExportBuilder.strip_if_string(extras.get("license_new"))), # required-if-applicable
("modified",
JsonExportBuilder.strip_if_string(extras.get("modified", package.get("metadata_modified")))),
# required
("primaryITInvestmentUII", JsonExportBuilder.strip_if_string(extras.get('primary_it_investment_uii'))),
# optional
# ('publisher', OrderedDict([
# ("@type", "org:Organization"),
# ("name", "Widget Services")
# ])), # required
# ("publisher", get_publisher_tree(extras)), # required
("publisher", JsonExportBuilder.get_publisher_tree_wrong_order(extras)), # required
("rights", JsonExportBuilder.strip_if_string(extras.get('access_level_comment'))), # required
("spatial", JsonExportBuilder.strip_if_string(package.get("spatial"))), # required-if-applicable
('systemOfRecords', JsonExportBuilder.strip_if_string(extras.get('system_of_records'))), # optional
("temporal", JsonExportBuilder.strip_if_string(extras.get('temporal'))), # required-if-applicable
("distribution", JsonExportBuilder.generate_distribution(package)), # required-if-applicable
# ("distribution",
# #TODO distribution should hide any key/value pairs where value is "" or None (e.g. format)
# [
# OrderedDict([
# ("downloadURL", r["url"]),
# ("mediaType", r["formatReadable"]),
# ])
# for r in package["resources"]
# ])
]
for pair in [
('bureauCode', 'bureau_code'), # required
('language', 'language'), # optional
('programCode', 'program_code'), # required
('references', 'related_documents'), # optional
('theme', 'category'), # optional
]:
JsonExportBuilder.split_multiple_entries(retlist, extras, pair)
except KeyError as e:
log.warn("Missing Required Field for package with id=[%s], title=['%s'], organization=['%s']: '%s'" % (
package.get('id'), package.get('title'), currentPackageOrg, e))
errors = ['Missing Required Field', ["%s" % e]]
errors_dict = OrderedDict([
('id', package.get('id')),
('name', package.get('name')),
('title', package.get('title')),
('organization', currentPackageOrg),
('errors', errors),
])
return errors_dict
# Remove entries where value is None, "", or empty list []
striped_retlist = [(x, y) for x, y in retlist if y is not None and y != "" and y != []]
# When saved from UI DataQuality value is stored as "on" instead of True.
# Check if value is "on" and replace it with True.
striped_retlist_dict = OrderedDict(striped_retlist)
if striped_retlist_dict.get('dataQuality') == "on" \
or striped_retlist_dict.get('dataQuality') == "true" \
or striped_retlist_dict.get('dataQuality') == "True":
striped_retlist_dict['dataQuality'] = True
elif striped_retlist_dict.get('dataQuality') == "false" \
or striped_retlist_dict.get('dataQuality') == "False":
striped_retlist_dict['dataQuality'] = False
from datajsonvalidator import do_validation
errors = []
try:
do_validation([dict(striped_retlist_dict)], errors)
except Exception as e:
errors.append(("Internal Error", ["Something bad happened: " + unicode(e)]))
if len(errors) > 0:
for error in errors:
log.warn(error)
errors_dict = OrderedDict([
('id', package.get('id')),
('name', package.get('name')),
('title', package.get('title')),
('organization', currentPackageOrg),
('errors', errors),
])
return errors_dict
return striped_retlist_dict
def make_datajson_entry(package):
# extras is a list of dicts [{},{}, {}]. For each dict, extract the key, value entries into a new dict
extras = dict([(x['key'], x['value']) for x in package['extras']])
parent_dataset_id = extras.get('parent_dataset')
if parent_dataset_id:
parent = model.Package.get(parent_dataset_id)
parent_uid = parent.extras.col.target['unique_id'].value
if parent_uid:
parent_dataset_id = parent_uid
# if resource format is CSV then convert it to text/csv
# Resource format has to be in 'csv' format for automatic datastore push.
for r in package["resources"]:
if r["format"].lower() == "csv":
r["format"] = "text/csv"
if r["format"].lower() == "json":
r["format"] = "application/json"
if r["format"].lower() == "pdf":
r["format"] = "application/pdf"
try:
retlist = [
("@type", "dcat:Dataset"), # optional
("title", strip_if_string(package["title"])), # required
# ("accessLevel", 'public'), # required
("accessLevel", strip_if_string(extras.get('public_access_level'))), # required
# ("accrualPeriodicity", "R/P1Y"), # optional
# ('accrualPeriodicity', 'accrual_periodicity'),
('accrualPeriodicity', get_accrual_periodicity(extras.get('accrual_periodicity'))), # optional
("conformsTo", strip_if_string(extras.get('conforms_to'))), # optional
# ('contactPoint', OrderedDict([
# ("@type", "vcard:Contact"),
# ("fn", "Jane Doe"),
# ("hasEmail", "mailto:[email protected]")
# ])), # required
('contactPoint', get_contact_point(extras, package)), # required
("dataQuality", strip_if_string(extras.get('data_quality'))), # required-if-applicable
("describedBy", strip_if_string(extras.get('data_dictionary'))), # optional
("describedByType", strip_if_string(extras.get('data_dictionary_type'))), # optional
("description", strip_if_string(package["notes"])), # required
# ("description", 'asdfasdf'), # required
("identifier", strip_if_string(extras.get('unique_id'))), # required
# ("identifier", 'asdfasdfasdf'), # required
("isPartOf", parent_dataset_id), # optional
("issued", strip_if_string(extras.get('release_date'))), # optional
# ("keyword", ['a', 'b']), # required
("keyword", [t["display_name"] for t in package["tags"]]), # required
("landingPage", strip_if_string(extras.get('homepage_url'))), # optional
("license", strip_if_string(extras.get("license_new"))), # required-if-applicable
("modified", strip_if_string(extras.get("modified"))), # required
("primaryITInvestmentUII", strip_if_string(extras.get('primary_it_investment_uii'))), # optional
# ('publisher', OrderedDict([
# ("@type", "org:Organization"),
# ("name", "Widget Services")
# ])), # required
("publisher", get_publisher_tree(extras)), # required
("rights", strip_if_string(extras.get('access_level_comment'))), # required
("spatial", strip_if_string(package.get("spatial"))), # required-if-applicable
('systemOfRecords', strip_if_string(extras.get('system_of_records'))), # optional
("temporal", strip_if_string(extras.get('temporal'))), # required-if-applicable
("distribution", generate_distribution(package)), # required-if-applicable
# ("distribution",
# #TODO distribution should hide any key/value pairs where value is "" or None (e.g. format)
# [
# OrderedDict([
# ("downloadURL", r["url"]),
# ("mediaType", r["formatReadable"]),
# ])
# for r in package["resources"]
# ])
]
for pair in [
('bureauCode', 'bureau_code'), # required
('language', 'language'), # optional
('programCode', 'program_code'), # required
('references', 'related_documents'), # optional
('theme', 'category'), # optional
]:
split_multiple_entries(retlist, extras, pair)
except KeyError as e:
log.warn("Invalid field detected for package with id=[%s], title=['%s']: '%s'", package.get('id'),
package.get('title'), e)
return
# # TODO this is a lazy hack to make sure we don't have redundant fields when the free form key/value pairs are added
# extras_to_filter_out = ['publisher', 'contact_name', 'contact_email', 'unique_id', 'public_access_level',
# 'data_dictionary', 'bureau_code', 'program_code', 'access_level_comment', 'license_title',
# 'spatial', 'temporal', 'release_date', 'accrual_periodicity', 'language', 'granularity',
# 'data_quality', 'size', 'homepage_url', 'rss_feed', 'category', 'related_documents',
# 'system_of_records', 'system_of_records_none_related_to_this_dataset', 'tags',
# 'extrasRollup', 'format', 'accessURL', 'notes', 'publisher_1', 'publisher_2', 'publisher_3',
# 'publisher_4', 'publisher_5']
#
# # Append any free extras (key/value pairs) that aren't part of common core but have been associated with the dataset
# # TODO really hackey, short on time, had to hardcode a lot of the names to remove. there's much better ways, maybe
# # generate a list of keys to ignore by calling a specific function to get the extras
# retlist_keys = [x for x, y in retlist]
# extras_keys = set(extras.keys()) - set(extras_to_filter_out)
#
# for key in extras_keys:
# convertedKey = underscore_to_camelcase(key)
# if convertedKey not in retlist_keys:
# retlist.append((convertedKey, extras[key]))
# Remove entries where value is None, "", or empty list []
striped_retlist = [(x, y) for x, y in retlist if y is not None and y != "" and y != []]
striped_retlist_keys = [x for x, y in striped_retlist]
# If a required metadata field was removed, return empty string
# for required_field in ["accessLevel", "bureauCode", "contactPoint", "description", "identifier", "keyword",
# "modified", "programCode", "publisher", "title"]:
# if required_field not in striped_retlist_keys:
# log.warn("Missing required field detected for package with id=[%s], title=['%s']: '%s'",
# package.get('id'), package.get('title'), required_field)
# return
# When saved from UI DataQuality value is stored as "on" instead of True.
# Check if value is "on" and replace it with True.
striped_retlist_dict = OrderedDict(striped_retlist)
if striped_retlist_dict.get('dataQuality') == "on" \
or striped_retlist_dict.get('dataQuality') == "true" \
or striped_retlist_dict.get('dataQuality') == "True":
striped_retlist_dict['dataQuality'] = True
elif striped_retlist_dict.get('dataQuality') == "false" \
or striped_retlist_dict.get('dataQuality') == "False":
striped_retlist_dict['dataQuality'] = False
from datajsonvalidator import do_validation
errors = []
try:
do_validation([dict(striped_retlist_dict)], errors)
except Exception as e:
errors.append(("Internal Error", ["Something bad happened: " + unicode(e)]))
if len(errors) > 0:
for error in errors:
log.warn(error)
return
return striped_retlist_dict
