class FunderAgent(Parser):
schema = tools.GuessAgentType(tools.OneOf(ctx.funderName,
ctx.contributorName),
default='organization')
name = tools.OneOf(ctx.funderName, ctx.contributorName)
identifiers = tools.Map(
tools.Delegate(AgentIdentifier),
tools.Try(tools.IRI(
tools.OneOf(ctx.funderIdentifier,
tools.RunPython(force_text, ctx.nameIdentifier),
tools.Static(None))),
exceptions=(ValueError, )))
class Extra:
name_identifier = tools.Try(ctx.nameIdentifier)
name_identifier_scheme = tools.Try(
ctx.nameIdentifier['@nameIdentifierScheme'])
name_identifier_scheme_uri = tools.Try(
ctx.nameIdentifier['@schemeURI'])
funder_identifier = tools.Try(ctx.funderIdentifier)
funder_identifier_type = tools.Try(ctx.funderIdentifierType)
contributor_type = tools.Try(ctx.contributorType)
class FundingAgent(Parser):
schema = tools.GuessAgentType(ctx.sponsorName, default='organization')
name = ctx.sponsorName
identifiers = tools.Map(tools.Delegate(AgentIdentifier),
tools.IRI(tools.Try(ctx.sponsorIdentifier)))
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 RelatedWorkIdentifier(Parser):
schema = 'WorkIdentifier'
uri = tools.IRI(tools.RunPython(force_text, ctx))
class Extra:
related_identifier_type = ctx['@relatedIdentifierType']
relation_type = tools.Try(ctx['@relationType'])
related_metadata_scheme = tools.Try(ctx['@relatedMetadataScheme'])
scheme_URI = tools.Try(ctx['@schemeURI'])
scheme_type = tools.Try(ctx['@schemeType'])
class Agent(Parser):
schema = tools.GuessAgentType(ctx.name)
name = ctx.name
related_agents = tools.Map(tools.Delegate(IsAffiliatedWith),
tools.Try(ctx.affiliation))
identifiers = tools.Map(
tools.Delegate(AgentIdentifier),
tools.Map(tools.IRI(), tools.Try(ctx.sameAs), tools.Try(ctx.email)))
class Extra:
givenName = tools.Try(ctx.givenName)
familyName = tools.Try(ctx.familyName)
additonalName = tools.Try(ctx.additionalName)
name = tools.Try(ctx.name)
class ContributorAgent(Parser):
schema = tools.OneOf(
tools.GuessAgentType(tools.RunPython(get_agent_type, ctx,
person=False),
default='organization'),
tools.GuessAgentType(tools.OneOf(ctx.creatorName,
ctx.contributorName)))
name = tools.OneOf(ctx.creatorName, ctx.contributorName)
identifiers = tools.Map(
tools.Delegate(AgentIdentifier),
tools.Try(tools.IRI(tools.RunPython(force_text, ctx.nameIdentifier)),
exceptions=(ValueError, )))
related_agents = tools.Map(
tools.Delegate(IsAffiliatedWith),
tools.Concat(
tools.Try(
tools.Filter(lambda x: bool(x),
tools.RunPython(force_text, ctx.affiliation)))))
class Extra:
name_identifier = tools.Try(ctx.nameIdentifier)
name_identifier_scheme = tools.Try(
ctx.nameIdentifier['@nameIdentifierScheme'])
name_identifier_scheme_uri = tools.Try(
ctx.nameIdentifier['@schemeURI'])
contributor_type = tools.Try(ctx.contributorType)
# v.4 new givenName and familyName properties
given_name = tools.OneOf(ctx.creatorName['@givenName'],
ctx.contributorName['@givenName'],
tools.Static(None))
family_name = tools.OneOf(ctx.creatorName['@familyName'],
ctx.contributorName['@familyName'],
tools.Static(None))
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 AgentIdentifier(Parser):
uri = tools.IRI(ctx)
class WorkIdentifier(Parser):
uri = tools.IRI(ctx)
class WorkIdentifier(Parser):
uri = tools.IRI(ctx['#text'])
class Award(Parser):
name = ctx.awardName
uri = tools.IRI(tools.Try(ctx.awardIdentifier))