class CreativeWork(Parser):
schema = tools.RunPython('get_type', ctx)
title = tools.RunPython('get_title', ctx)
description = Soup(ctx, 'p', class_='genericfile_description')['#text']
date_published = tools.ParseDate(
Soup(ctx, itemprop='datePublished')['#text'])
date_updated = tools.ParseDate(Soup(ctx, itemprop='dateModified')['#text'])
rights = tools.OneOf(tools.RunPython('get_rights_url', ctx),
tools.RunPython('get_dd', ctx, 'Rights')['#text'],
tools.Static(None))
language = tools.Try(
tools.ParseLanguage(Soup(ctx, itemprop='inLanguage')['#text']))
tags = tools.Map(tools.Delegate(ThroughTags), Soup(ctx,
itemprop='keywords'))
identifiers = tools.Map(
tools.Delegate(WorkIdentifier),
tools.Try(tools.RunPython('get_dd', ctx, 'Permanent Link')),
)
related_agents = tools.Concat(
tools.Map(tools.Delegate(Creator), Soup(ctx, itemprop='creator')),
tools.Map(tools.Delegate(Contributor), Soup(ctx,
itemprop='contributor')),
tools.Map(tools.Delegate(Publisher), Soup(ctx, itemprop='publisher')),
)
class Extra:
gwu_unit = tools.RunPython('get_dd', ctx, 'GW Unit')['#text']
related_url = tools.RunPython('get_dd', ctx, 'Related URL')['#text']
previous_publication_information = tools.RunPython(
'get_dd', ctx, 'Previous Publication Information')['#text']
depositor = tools.RunPython('get_dd', ctx, 'Depositor')['#text']
characterization = tools.RunPython('get_dd', ctx,
'Characterization')['#text']
def get_type(self, obj):
return {
'http://schema.org/CreativeWork': 'CreativeWork',
'http://schema.org/Article': 'Article',
'http://schema.org/Book': 'Book',
}.get(obj.soup.find('div')['itemtype'], 'CreativeWork')
def get_title(self, obj):
title = obj.h1.soup
title.find('span', class_='label').decompose()
return title.get_text()
def get_dd(self, obj, dt):
dt_tag = obj.soup.find('dt', string=dt)
if dt_tag:
return SoupXMLDict(soup=dt_tag.find_next_sibling('dd'))
return None
def get_rights_url(self, obj):
dd = self.get_dd(obj, 'Rights')
return dd.soup.find('i', class_='glyphicon-new-window').parent['href']
class CreativeWork(Parser):
title = ctx.title
description = tools.Try(ctx.description)
is_deleted = tools.RunPython('_is_deleted', tools.Try(ctx.otherProperties))
date_updated = tools.ParseDate(tools.Try(ctx.providerUpdatedDateTime))
rights = tools.Join(tools.Try(ctx.licenses.uri))
# Note: this is only taking the first language in the case of multiple languages
language = tools.ParseLanguage(tools.Try(ctx.languages[0]), )
related_agents = tools.Concat(
tools.Map(tools.Delegate(Creator), tools.Try(ctx.contributors)),
tools.Map(tools.Delegate(Publisher), tools.Try(ctx.publisher)),
tools.Map(tools.Delegate(Funder), tools.Try(ctx.sponsorships)))
identifiers = tools.Map(
tools.Delegate(WorkIdentifier),
tools.Map(
tools.IRI(),
tools.RunPython(
'unique',
tools.Concat(tools.Try(ctx.uris.canonicalUri),
tools.Try(ctx.uris.providerUris),
tools.Try(ctx.uris.descriptorUris),
tools.Try(ctx.uris.objectUris)))))
subjects = tools.Map(tools.Delegate(ThroughSubjects),
tools.Subjects(tools.Try(ctx.subjects)))
tags = tools.Map(tools.Delegate(ThroughTags), tools.Try(ctx.tags),
tools.Try(ctx.subjects))
class Extra:
"""
Fields that are combined in the base parser are relisted as singular elements that match
their original entry to preserve raw data structure.
"""
freeToRead = tools.Try(ctx.freeToRead)
languages = tools.Try(ctx.languages)
licenses = tools.Try(ctx.licenses)
otherProperties = tools.Try(ctx.otherProperties)
publisher = tools.Try(ctx.publisher)
subjects = tools.Try(ctx.subjects)
sponsorships = tools.Try(ctx.sponsorships)
tags = tools.Try(ctx.tags)
uris = tools.Try(ctx.uris)
version = tools.Try(ctx.version)
def unique(self, items):
return list(sorted(set(items)))
def _is_deleted(self, properties):
for prop in properties or []:
if prop['name'] == 'status':
return 'deleted' in prop['properties'].get('status', [])
return False
class OAICreativeWork(Parser):
default_type = None
type_map = None
schema = tools.RunPython(
'get_schema',
tools.OneOf(ctx.record.metadata.dc['dc:type'], tools.Static(None)))
title = tools.Join(
tools.RunPython('force_text',
tools.Try(ctx.record.metadata.dc['dc:title'])))
description = tools.Join(
tools.RunPython('force_text',
tools.Try(ctx.record.metadata.dc['dc:description'])))
identifiers = tools.Map(
tools.Delegate(OAIWorkIdentifier),
tools.Unique(
tools.Map(
tools.Try(tools.IRI(), exceptions=(ValueError, )),
tools.Filter(
not_citation,
tools.RunPython(
'force_text',
tools.Concat(
tools.Try(ctx.record.metadata.dc['dc:identifier']),
tools.Try(ctx.record.header['identifier'])))))))
related_works = tools.Concat(
tools.Map(
tools.Delegate(OAIWorkRelation),
tools.Unique(
tools.Map(tools.Try(tools.IRI(), exceptions=(ValueError, )),
tools.RunPython('get_relation', ctx)))))
related_agents = tools.Concat(
tools.Map(tools.Delegate(OAICreator),
tools.Try(ctx.record.metadata.dc['dc:creator'])),
tools.Map(tools.Delegate(OAIContributor),
tools.Try(ctx.record.metadata.dc['dc:contributor'])),
tools.Map(
tools.Delegate(OAIPublisher),
tools.RunPython('force_text',
tools.Try(
ctx.record.metadata.dc['dc:publisher']))),
)
rights = tools.Join(tools.Try(ctx.record.metadata.dc['dc:rights']))
# Note: this is only taking the first language in the case of multiple languages
language = tools.ParseLanguage(
tools.Try(ctx.record.metadata.dc['dc:language'][0]), )
subjects = tools.Map(
tools.Delegate(OAIThroughSubjects),
tools.Subjects(
tools.Map(
tools.RunPython('tokenize'),
tools.RunPython(
'force_text',
tools.Concat(
tools.Try(ctx.record.header.setSpec),
tools.Try(ctx.record.metadata.dc['dc:type']),
tools.Try(ctx.record.metadata.dc['dc:format']),
tools.Try(ctx.record.metadata.dc['dc:subject']),
)))))
tags = tools.Map(
tools.Delegate(OAIThroughTags),
tools.Concat(tools.Map(
tools.RunPython('tokenize'),
tools.RunPython(
'force_text',
tools.Concat(
tools.Try(ctx.record.header.setSpec),
tools.Try(ctx.record.metadata.dc['dc:type']),
tools.Try(ctx.record.metadata.dc['dc:format']),
tools.Try(ctx.record.metadata.dc['dc:subject']),
))),
deep=True))
date_updated = tools.ParseDate(ctx.record.header.datestamp)
is_deleted = tools.RunPython('check_status',
tools.Try(ctx.record.header['@status']))
class Extra:
"""
Fields that are combined in the base parser are relisted as singular elements that match
their original entry to preserve raw data structure.
"""
# An agent responsible for making contributions to the resource.
contributor = tools.Try(ctx.record.metadata.dc['dc:contributor'])
# The spatial or temporal topic of the resource, the spatial applicability of the resource,
# or the jurisdiction under which the resource is relevant.
coverage = tools.Try(ctx.record.metadata.dc['dc:coverage'])
# An agent primarily responsible for making the resource.
creator = tools.Try(ctx.record.metadata.dc['dc:creator'])
# A point or period of time associated with an event in the lifecycle of the resource.
dates = tools.Try(ctx.record.metadata.dc['dc:date'])
# The file format, physical medium, or dimensions of the resource.
resource_format = tools.Try(ctx.record.metadata.dc['dc:format'])
# An unambiguous reference to the resource within a given context.
identifiers = tools.Concat(
tools.Try(ctx.record.metadata.dc['dc:identifier']),
tools.Try(ctx.record.header['identifier']))
# A related resource.
relation = tools.RunPython('get_relation', ctx)
# A related resource from which the described resource is derived.
source = tools.Try(ctx.record.metadata.dc['dc:source'])
# The nature or genre of the resource.
resource_type = tools.Try(ctx.record.metadata.dc['dc:type'])
set_spec = tools.Try(ctx.record.header.setSpec)
# Language also stored in the Extra class in case the language reported cannot be parsed by ParseLanguage
language = tools.Try(ctx.record.metadata.dc['dc:language'])
# Status in the header, will exist if the resource is deleted
status = tools.Try(ctx.record.header['@status'])
def check_status(self, status):
if status == 'deleted':
return True
return False
def get_schema(self, types):
if not types or not self.type_map:
return self.default_type
if isinstance(types, str):
types = [types]
for t in types:
if isinstance(t, dict):
t = t['#text']
t = t.lower()
if t in self.type_map:
return self.type_map[t]
return self.default_type
def force_text(self, data):
if isinstance(data, dict):
return data['#text']
if isinstance(data, str):
return data
fixed = []
for datum in (data or []):
if datum is None:
continue
if isinstance(datum, dict):
if '#text' not in datum:
logger.warn('Skipping %s, no #text key exists', datum)
continue
fixed.append(datum['#text'])
elif isinstance(datum, str):
fixed.append(datum)
else:
raise Exception(datum)
return fixed
def tokenize(self, data):
if isinstance(data, str):
data = [data]
tokens = []
for item in data:
tokens.extend(
[x.strip() for x in re.split(r'(?: - )|\.|,', item) if x])
return tokens
def get_relation(self, ctx):
if not ctx['record'].get('metadata'):
return []
relation = ctx['record']['metadata']['dc'].get('dc:relation') or []
identifiers = ctx['record']['metadata']['dc'].get(
'dc:identifier') or []
if isinstance(identifiers, dict):
identifiers = (identifiers, )
identifiers = ''.join(i['#text'] if isinstance(i, dict) else i
for i in identifiers if i)
identifiers = re.sub(
'http|:|/', '',
identifiers + ctx['record']['header']['identifier'])
if isinstance(relation, dict):
relation = (relation['#text'], )
return [
r for r in relation
if r and re.sub('http|:|/', '', r) not in identifiers
]
class OAICreativeWork(Parser):
schema = 'CreativeWork'
ORGANIZATION_KEYWORDS = ('the', 'center')
INSTITUTION_KEYWORDS = ('school', 'university', 'institution', 'institute')
title = tools.Join(
tools.RunPython('force_text',
tools.Try(
ctx['record']['metadata']['dc']['dc:title'])))
description = tools.Join(
tools.RunPython('force_text',
tools.Try(ctx.record.metadata.dc['dc:description'])))
publishers = tools.Map(
tools.Delegate(
OAIAssociation.using(entity=tools.Delegate(OAIPublisher))),
tools.Map(tools.RunPython('force_text'),
tools.Try(ctx.record.metadata.dc['dc:publisher'])))
rights = tools.Join(
tools.Maybe(tools.Maybe(ctx['record'], 'metadata')['dc'], 'dc:rights'))
# Note: this is only taking the first language in the case of multiple languages
language = tools.ParseLanguage(
tools.Try(ctx['record']['metadata']['dc']['dc:language'][0]), )
contributors = tools.Map(
tools.Delegate(OAIContributor),
tools.RunPython(
'get_contributors',
tools.Concat(
tools.Maybe(
tools.Maybe(ctx['record'], 'metadata')['dc'],
'dc:creator'),
tools.Maybe(
tools.Maybe(ctx['record'], 'metadata')['dc'],
'dc:contributor')), 'contributor'))
institutions = tools.Map(
tools.Delegate(
OAIAssociation.using(entity=tools.Delegate(OAIInstitution))),
tools.RunPython(
'get_contributors',
tools.Concat(
tools.Maybe(
tools.Maybe(ctx['record'], 'metadata')['dc'],
'dc:creator'),
tools.Maybe(
tools.Maybe(ctx['record'], 'metadata')['dc'],
'dc:contributor')), 'institution'))
organizations = tools.Map(
tools.Delegate(
OAIAssociation.using(entity=tools.Delegate(OAIOrganization))),
tools.RunPython(
'get_contributors',
tools.Concat(
tools.Maybe(
tools.Maybe(ctx['record'], 'metadata')['dc'],
'dc:creator'),
tools.Maybe(
tools.Maybe(ctx['record'], 'metadata')['dc'],
'dc:contributor')), 'organization'))
tags = tools.Map(
tools.Delegate(OAIThroughTags),
tools.RunPython(
'force_text',
tools.Concat(
tools.Try(ctx['record']['header']['setSpec']),
tools.Try(ctx['record']['metadata']['dc']['dc:type']),
tools.Try(ctx['record']['metadata']['dc']['dc:format']),
tools.Try(ctx['record']['metadata']['dc']['dc:subject']),
)))
links = tools.Map(
tools.Delegate(OAIThroughLinks),
tools.RunPython(
'get_links',
tools.Concat(
tools.Try(ctx['record']['metadata']['dc']['dc:identifier']),
tools.Maybe(
tools.Maybe(ctx['record'], 'metadata')['dc'],
'dc:relation'))))
date_updated = tools.ParseDate(ctx['record']['header']['datestamp'])
class Extra:
"""
Fields that are combined in the base parser are relisted as singular elements that match
their original entry to preserve raw data structure.
"""
# An entity responsible for making contributions to the resource.
contributor = tools.Maybe(
tools.Maybe(ctx['record'], 'metadata')['dc'], 'dc:contributor')
# The spatial or temporal topic of the resource, the spatial applicability of the resource,
# or the jurisdiction under which the resource is relevant.
coverage = tools.Maybe(
tools.Maybe(ctx['record'], 'metadata')['dc'], 'dc:coverage')
# An entity primarily responsible for making the resource.
creator = tools.Maybe(
tools.Maybe(ctx['record'], 'metadata')['dc'], 'dc:creator')
# A point or period of time associated with an event in the lifecycle of the resource.
dates = tools.Maybe(
tools.Maybe(ctx['record'], 'metadata')['dc'], 'dc:date')
# The file format, physical medium, or dimensions of the resource.
resource_format = tools.Maybe(
tools.Maybe(ctx['record'], 'metadata')['dc'], 'dc:format')
# An unambiguous reference to the resource within a given context.
identifiers = tools.Concat(
tools.Try(ctx['record']['metadata']['dc']['dc:identifier']),
tools.Maybe(ctx['record']['header'], 'identifier'))
# A related resource.
relation = tools.RunPython('get_relation', ctx)
# A related resource from which the described resource is derived.
source = tools.Maybe(
tools.Maybe(ctx['record'], 'metadata')['dc'], 'dc:source')
# The topic of the resource.
subject = tools.Try(ctx.record.metadata.dc['dc:subject'])
# The nature or genre of the resource.
resource_type = tools.Try(ctx.record.metadata.dc['dc:type'])
set_spec = tools.Maybe(ctx.record.header, 'setSpec')
# Language also stored in the Extra class in case the language reported cannot be parsed by ParseLanguage
language = tools.Try(ctx.record.metadata.dc['dc:language'])
# Status in the header, will exist if the resource is deleted
status = tools.Maybe(ctx.record.header, '@status')
def get_links(self, ctx):
links = []
for link in ctx:
if not link or not isinstance(link, str):
continue
found_url = URL_REGEX.search(link)
if found_url is not None:
links.append(found_url.group())
continue
found_doi = DOI_REGEX.search(link)
if found_doi is not None:
found_doi = found_doi.group()
if 'dx.doi.org' in found_doi:
links.append(found_doi)
else:
links.append('http://dx.doi.org/{}'.format(
found_doi.replace('doi:', '')))
return links
def force_text(self, data):
if isinstance(data, dict):
return data['#text']
if isinstance(data, str):
return data
fixed = []
for datum in (data or []):
if datum is None:
continue
if isinstance(datum, dict):
if '#text' not in datum:
logger.warn('Skipping %s, no #text key exists', datum)
continue
fixed.append(datum['#text'])
elif isinstance(datum, str):
fixed.append(datum)
else:
raise Exception(datum)
return fixed
def get_relation(self, ctx):
if not ctx['record'].get('metadata'):
return []
relation = ctx['record']['metadata']['dc'].get('dc:relation', [])
if isinstance(relation, dict):
return relation['#text']
return relation
def get_contributors(self, options, entity):
"""
Returns list of organization, institutions, or contributors names based on entity type.
"""
options = [o if isinstance(o, str) else o['#text'] for o in options]
if entity == 'organization':
organizations = [
value for value in options
if (value and
not self.list_in_string(value, self.INSTITUTION_KEYWORDS)
and self.list_in_string(value, self.ORGANIZATION_KEYWORDS))
]
return organizations
elif entity == 'institution':
institutions = [
value for value in options
if (value
and self.list_in_string(value, self.INSTITUTION_KEYWORDS))
]
return institutions
elif entity == 'contributor':
people = [
value for value in options
if (value and not self.list_in_string(
value, self.INSTITUTION_KEYWORDS) and
not self.list_in_string(value, self.ORGANIZATION_KEYWORDS))
]
return people
else:
return options
def list_in_string(self, string, list_):
if any(word in string.lower() for word in list_):
return True
return False
class CreativeWork(Parser):
# Schema definitions: http://schema.datacite.org/meta/kernel-3.1/doc/DataCite-MetadataKernel_v3.1.pdf
PEOPLE_TYPES = ('ContactPerson', 'DataCurator', 'Editor', 'ProjectLeader',
'ProjectManager', 'ProjectMember', 'RelatedPerson',
'Researcher', 'Supervisor', 'WorkPackageLeader')
CHECK_TYPES = ('DataCollector', 'DataManager', 'Producer', 'RightsHolder',
'Sponsor', 'Other')
NOT_PEOPLE_TYPES = ('Distributor', 'HostingInstitution',
'RegistrationAgency', 'RegistrationAuthority',
'ResearchGroup')
ORGANIZATION_KEYWORDS = (THE_REGEX, 'council', 'center', 'foundation')
INSTITUTION_KEYWORDS = ('school', 'university', 'institution', 'college',
'institute')
title = tools.RunPython(
force_text,
tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.titles.title))
description = tools.RunPython(
force_text,
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
descriptions.description[0]))
publishers = tools.Map(
tools.Delegate(Association.using(entity=tools.Delegate(Publisher))),
tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.publisher))
rights = tools.Try(
tools.Join(
tools.RunPython(
'text_list',
tools.Concat(ctx.record.metadata['oai_datacite'].payload.
resource.rightsList.rights))))
language = tools.ParseLanguage(
tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.language))
contributors = tools.Concat(
tools.Map(
tools.Delegate(Creator),
tools.RunPython(
'get_contributors',
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.creators.creator)), 'contributor',
'creator')),
tools.Map(
tools.Delegate(Contributor),
tools.RunPython(
'get_contributors',
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.contributors.contributor)),
'contributor', 'contributor')))
institutions = tools.Concat(
tools.Map(
tools.Delegate(
Association.using(entity=tools.Delegate(CreatorInstitution))),
tools.RunPython(
'get_contributors',
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.creators.creator)), 'institution',
'creator')),
tools.Map(
tools.Delegate(
Association.using(
entity=tools.Delegate(ContributorInstitution))),
tools.RunPython(
'get_contributors',
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.contributors.contributor)),
'institution', 'contributor')))
organizations = tools.Concat(
tools.Map(
tools.Delegate(
Association.using(entity=tools.Delegate(CreatorOrganization))),
tools.RunPython(
'get_contributors',
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.creators.creator)), 'organization',
'creator')),
tools.Map(
tools.Delegate(
Association.using(
entity=tools.Delegate(ContributorOrganization))),
tools.RunPython(
'get_contributors',
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.contributors.contributor)),
'organization', 'contributor')))
tags = tools.Map(
tools.Delegate(ThroughTags),
tools.RunPython(
force_text,
tools.Concat(
tools.Maybe(
tools.Maybe(ctx.record, 'metadata')['oai_datacite'],
'type'),
tools.RunPython('text_list', (tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.subjects.subject)))),
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
formats.format),
tools.Try(
ctx.record.metadata['oai_datacite'].datacentreSymbol),
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
resourceType['#text']),
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
resourceType['@resourceTypeGeneral']),
tools.Maybe(ctx.record.header, 'setSpec'),
tools.Maybe(ctx.record.header, '@status'))))
links = tools.Concat(
tools.Map(
tools.Delegate(ThroughLinks),
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
identifier))),
tools.Map(
tools.Delegate(ThroughAlternateLinks),
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
alternateIdentifiers.alternateidentifier))),
tools.Map(
tools.Delegate(ThroughRelatedLinks),
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
relatedIdentifiers.relatedIdentifier))))
date_updated = tools.ParseDate(ctx.record.header.datestamp)
date_published = tools.ParseDate(
tools.RunPython(
'get_date_type',
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
dates.date)), 'Issued'))
free_to_read_type = tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.rightsList.rights['@rightsURI'])
free_to_read_date = tools.ParseDate(
tools.RunPython(
'get_date_type',
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
dates.date)), 'Available'))
funders = tools.Map(
tools.Delegate(Association.using(entity=tools.Delegate(Funder))),
tools.RunPython(
'get_contributors',
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
contributors.contributor)), 'funder', 'contributor'))
venues = tools.Map(
tools.Delegate(ThroughVenues),
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
geoLocations.geoLocation))
class Extra:
"""
Fields that are combined in the base parser are relisted as singular elements that match
their original entry to preserve raw data structure.
"""
status = tools.Maybe(ctx.record.header, '@status')
datestamp = tools.ParseDate(ctx.record.header.datestamp)
set_spec = tools.Maybe(ctx.record.header, 'setSpec')
is_reference_quality = tools.Try(
ctx.record.metadata['oai_datacite'].isReferenceQuality)
schema_version = tools.Try(
ctx.record.metadata['oai_datacite'].schemaVersion)
datacentre_symbol = tools.Try(
ctx.record.metadata['oai_datacite'].datacentreSymbol)
identifiers = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.identifier)
alternate_identifiers = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.
alternateIdentifiers.alternateidentifier)
titles = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.titles.title)
publisher = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.publisher)
publication_year = tools.Try(ctx.record.metadata['oai_datacite'].
payload.resource.publicationYear)
subject = tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.subjects.subject)
resourceType = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.resourceType)
sizes = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.size)
format_type = tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.formats.format)
version = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.version)
rights = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.rights)
rightsList = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.rightsList)
related_identifiers = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.
relatedIdentifiers.relatedIdentifier)
description = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.descriptions)
dates = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.dates.date)
contributors = tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.contributors.contributor)
creators = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.creators)
geolocations = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.geoLocations)
def get_date_type(self, date_obj, date_type):
try:
date = None
for obj in date_obj:
if obj['@dateType'] == date_type:
date = obj['#text']
if date and not DATE_REGEX.search(date):
return None
except KeyError:
return None
return date
def text_list(self, data):
text_list = []
if isinstance(data, list):
for item in data:
if isinstance(item, dict):
if '#text' in item:
text_list.append(item['#text'])
continue
elif isinstance(item, str):
text_list.append(item)
continue
logger.warning(
'#text is not in {} and it is not a string'.format(item))
return text_list
else:
raise Exception('{} is not a list.'.format(data))
def get_contributors(self, options, entity, field=None):
"""
Returns list of organization, institutions, or contributors names based on entity type.
"""
if entity == 'organization':
organizations = []
for value in options:
val = self.try_contributor_type(value, self.NOT_PEOPLE_TYPES)
if val:
if field == 'creator':
organizations.append(val[field + 'Name'])
else:
organizations.append(val)
elif (value[field + 'Name'] and not self.list_in_string(
value[field + 'Name'], self.INSTITUTION_KEYWORDS)
and self.list_in_string(value[field + 'Name'],
self.ORGANIZATION_KEYWORDS)):
if field == 'creator':
organizations.append(value[field + 'Name'])
else:
organizations.append(value)
return organizations
elif entity == 'institution':
institutions = []
for value in options:
val = self.try_contributor_type(value, self.NOT_PEOPLE_TYPES)
if val:
institutions.append(val)
elif (value[field + 'Name'] and self.list_in_string(
value[field + 'Name'], self.INSTITUTION_KEYWORDS)):
institutions.append(value)
return institutions
elif entity == 'contributor':
people = []
for value in options:
val = self.try_contributor_type(value, self.PEOPLE_TYPES)
if val:
people.append(val)
elif (value[field + 'Name'] and not self.list_in_string(
value[field + 'Name'], self.INSTITUTION_KEYWORDS)
and not self.list_in_string(value[field + 'Name'],
self.ORGANIZATION_KEYWORDS)):
people.append(value)
return people
elif entity == 'funder':
funders = []
for value in options:
val = self.try_contributor_type(value, ['Funder'])
if val:
funders.append(val)
return funders
else:
return options
def try_contributor_type(self, value, target_list_types):
try:
contrib_type_item = value['@contributorType']
if contrib_type_item in target_list_types:
return value
return None
except KeyError:
return None
def list_in_string(self, string, list_):
for word in list_:
if isinstance(word, str):
if word in string.lower():
return True
else:
if word.search(string):
return True
return False
class CreativeWork(Parser):
'''
Documentation for Datacite's metadata:
https://schema.labs.datacite.org/meta/kernel-4.0/doc/DataCite-MetadataKernel_v4.0.pdf
'''
def get_schema(self, type):
return {
'dataset': 'DataSet',
'software': 'Software',
'text/book': 'Book',
'text/book chapter': 'Book',
'text/book prospectus': 'Book',
'text/book series': 'Book',
'text/conference abstract': 'ConferencePaper',
'text/conference paper': 'ConferencePaper',
'text/conference poster': 'Poster',
'text/dissertation': 'Dissertation',
'text/edited book': 'Book',
'text/journal article': 'Article',
'text/journal issue': 'Article',
'text/patent': 'Patent',
'text/report': 'Report',
'text/supervised student publication': 'Thesis',
'text/working paper': 'WorkingPaper'
# 'audiovisual': '',
# 'collection': '',
# 'event': '',
# 'image': '',
# 'interactiveresource': '',
# 'model': '',
# 'physicalobject': '',
# 'service': '',
# 'sound': '',
# 'text15': '',
# 'workflow': '',
# 'text/book review': '',
# 'text/conference program': '',
# 'text/dictionary entry': '',
# 'text/disclosure': '',
# 'text/encyclopedia entry': '',
# 'text/Funding submission': '',
# 'text/license': '',
# 'text/magazine article': '',
# 'text/manual': '',
# 'text/newsletter article': '',
# 'text/newspaper article': '',
# 'text/online resource': '',
# 'text/registered copyright': '',
# 'text/research tool': '',
# 'text/tenure-promotion': '',
# 'text/test': '',
# 'text/trademark': '',
# 'text/translation': '',
# 'text/university academic unit': '',
# 'text/website': '',
}.get(type.lower()) or 'CreativeWork'
schema = tools.RunPython(
'get_schema',
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
resourceType['@resourceTypeGeneral'],
default='CreativeWork'))
title = tools.RunPython(
force_text,
tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.titles.title))
description = tools.RunPython(
force_text,
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
descriptions.description[0]))
rights = tools.Try(
tools.Join(
tools.RunPython(
'text_list',
tools.Concat(ctx.record.metadata['oai_datacite'].payload.
resource.rightsList.rights))))
language = tools.ParseLanguage(
tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.language))
related_agents = tools.Concat(
tools.Map(
tools.Delegate(CreatorRelation),
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
creators.creator))),
tools.Map(
tools.Delegate(ContributorRelation),
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
contributors.contributor))),
tools.Map(
tools.Delegate(PublisherRelation),
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
publisher)),
tools.Map(
tools.Delegate(HostRelation),
tools.RunPython(
get_contributors,
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.contributors.contributor)),
['HostingInstitution'])),
# v.3 Funder is a contributor type
# v.4 FundingReference replaces funder contributor type
tools.Map(
tools.Delegate(FunderRelation),
tools.RunPython(
get_contributors,
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.contributors.contributor)),
['Funder'])),
tools.Map(
tools.Delegate(FunderRelation),
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
fundingReference))))
# v.4 New, free text, 'subjectScheme' attribute on subject
subjects = tools.Map(
tools.Delegate(ThroughSubjects),
tools.Subjects(
tools.RunPython(
'text_list',
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.subjects.subject), ))))
tags = tools.Map(
tools.Delegate(ThroughTags),
tools.RunPython(
force_text,
tools.Concat(
tools.Maybe(
tools.Maybe(ctx.record, 'metadata')['oai_datacite'],
'type'),
tools.RunPython('text_list', (tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.subjects.subject)))),
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
formats.format),
tools.Try(
ctx.record.metadata['oai_datacite'].datacentreSymbol),
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
resourceType['#text']),
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
resourceType['@resourceTypeGeneral']),
tools.Maybe(ctx.record.header, 'setSpec'),
tools.Maybe(ctx.record.header, '@status'))))
identifiers = tools.Concat(
tools.Map(
tools.Delegate(WorkIdentifier),
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
identifier))),
tools.Map(
tools.Delegate(WorkIdentifier),
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.resource.
alternateIdentifiers.alternateidentifier))))
related_works = tools.Concat(
tools.Map(
tools.Delegate(WorkRelation),
tools.RunPython(
get_related_works,
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.relatedIdentifiers.relatedIdentifier)),
False)),
tools.Map(
tools.Delegate(InverseWorkRelation),
tools.RunPython(
get_related_works,
tools.Concat(
tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.relatedIdentifiers.relatedIdentifier)),
True)))
date_updated = tools.ParseDate(tools.Try(ctx.record.header.datestamp))
date_published = tools.ParseDate(
tools.Try(
tools.RunPython(
'get_date_type',
tools.Concat(ctx.record.metadata['oai_datacite'].payload.
resource.dates.date), 'Issued')))
free_to_read_type = tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.rightsList.rights['@rightsURI'])
free_to_read_date = tools.ParseDate(
tools.Try(
tools.RunPython(
'get_date_type',
tools.Concat(ctx.record.metadata['oai_datacite'].payload.
resource.dates.date), 'Available')))
is_deleted = tools.RunPython('check_status',
tools.Try(ctx.record.header['@status']))
class Extra:
"""
Fields that are combined in the base parser are relisted as singular elements that match
their original entry to preserve raw data structure.
"""
status = tools.Try(ctx.record.header['@status'])
datestamp = tools.ParseDate(ctx.record.header.datestamp)
set_spec = tools.Try(ctx.record.header.setSpec)
is_reference_quality = tools.Try(
ctx.record.metadata['oai_datacite'].isReferenceQuality)
schema_version = tools.Try(
ctx.record.metadata['oai_datacite'].schemaVersion)
datacentre_symbol = tools.Try(
ctx.record.metadata['oai_datacite'].datacentreSymbol)
identifiers = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.identifier)
alternate_identifiers = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.
alternateIdentifiers.alternateidentifier)
titles = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.titles.title)
publisher = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.publisher)
publication_year = tools.Try(ctx.record.metadata['oai_datacite'].
payload.resource.publicationYear)
subject = tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.subjects.subject)
resourceType = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.resourceType)
sizes = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.size)
format_type = tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.formats.format)
version = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.version)
rights = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.rights)
rightsList = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.rightsList)
related_identifiers = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.
relatedIdentifiers.relatedIdentifier)
description = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.descriptions)
dates = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.dates.date)
contributors = tools.Try(ctx.record.metadata['oai_datacite'].payload.
resource.contributors.contributor)
creators = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.creators)
# v.4 new property geoLocationPolygon, in addition to geoLocationPoint and geoLocationBox
geolocations = tools.Try(
ctx.record.metadata['oai_datacite'].payload.resource.geoLocations)
funding_reference = tools.Try(ctx.record.metadata['oai_datacite'].
payload.resource.fundingReference)
def check_status(self, status):
if status == 'deleted':
return True
return False
def get_date_type(self, date_obj, date_type):
date = None
for obj in date_obj:
if obj['@dateType'] == date_type:
date = obj['#text']
if date and date != '0000':
return date
# raise KeyError to break TryLink
raise KeyError()
def text_list(self, data):
text_list = []
if isinstance(data, list):
for item in data:
if isinstance(item, dict):
if '#text' in item:
text_list.append(item['#text'])
continue
elif isinstance(item, str):
text_list.append(item)
continue
logger.warning(
'#text is not in {} and it is not a string'.format(item))
return text_list
else:
raise Exception('{} is not a list.'.format(data))