def resolve_viaf(self, work):
"""Get VIAF data on all contributors."""
viaf = VIAFClient(self._db)
for pool in work.license_pools:
edition = pool.presentation_edition
for contributor in edition.contributors:
viaf.process_contributor(contributor)
if not contributor.display_name:
contributor.family_name, contributor.display_name = (
contributor.default_names())
class RedoOCLCForThreeMScript(Script):
def __init__(self, test_session=None):
# Allows tests to run without db session overlap.
if test_session:
self._session = test_session
self.coverage = LinkedDataCoverageProvider(self._db)
self.oclc_classify = OCLCClassifyCoverageProvider(self._db)
self.viaf = VIAFClient(self._db)
@property
def input_data_source(self):
return DataSource.lookup(self._db, DataSource.OCLC_LINKED_DATA)
def do_run(self):
"""Re-runs OCLC Linked Data coverage provider to get viafs. Fetches
author information and recalculates presentation."""
identifiers = self.fetch_authorless_threem_identifiers()
self.delete_coverage_records(identifiers)
self.ensure_isbn_identifier(identifiers)
for identifier in identifiers:
self.coverage.ensure_coverage(identifier)
self.merge_contributors(identifier)
# Recalculate everything so the contributors can be seen.
for contributor in identifier.primary_edition.contributors:
self.viaf.process_contributor(contributor)
identifier.primary_edition.calculate_presentation()
if identifier.licensed_through:
identifier.licensed_through.calculate_work()
def fetch_authorless_threem_identifiers(self):
"""Returns a list of ThreeM identifiers that don't have contributors"""
qu = self._db.query(Identifier).join(Identifier.primarily_identifies)
qu = qu.outerjoin(
Edition.contributions).filter(Contribution.id == None)
qu = qu.filter(Identifier.type == Identifier.THREEM_ID)
return qu.all()
def delete_coverage_records(self, identifiers):
"""Deletes existing OCLC Linked Data coverage records to re-run and
capture author data"""
t1 = self._db.begin_nested()
for identifier in identifiers:
for coverage_record in identifier.coverage_records:
if coverage_record.data_source == self.input_data_source:
self._db.delete(coverage_record)
t1.commit()
def ensure_isbn_identifier(self, identifiers):
"""Runs OCLCClassify to get ISBN numbers if they're not available."""
identifiers_without_isbn = []
for identifier in identifiers:
equivalencies = identifier.equivalencies
equivalent_types = [eq.output.type for eq in equivalencies]
if Identifier.ISBN not in equivalent_types:
identifiers_without_isbn.append(identifier)
for identifier in identifiers_without_isbn:
self.oclc_classify.ensure_coverage(identifier)
def merge_contributors(self, identifier):
"""Gives a ThreeM primary edition any contributors found via OCLC-LD"""
qu = self._db.query(Identifier).join(Identifier.inbound_equivalencies)
qu = qu.filter(
or_(Identifier.type == Identifier.OCLC_WORK,
Identifier.type == Identifier.OCLC_NUMBER)).filter(
Equivalency.input_id == identifier.id)
oclc_contributions = []
for oclc_identifier in qu.all():
editions = oclc_identifier.primarily_identifies
for edition in editions:
oclc_contributions += edition.contributions
for contribution in oclc_contributions:
for edition in identifier.primarily_identifies:
edition.add_contributor(contribution.contributor,
contribution.role)
class RedoOCLCForThreeMScript(Script):
def __init__(self, test_session=None):
# Allows tests to run without db session overlap.
if test_session:
self._session = test_session
self.coverage = LinkedDataCoverageProvider(self._db)
self.oclc_classify = OCLCClassifyCoverageProvider(self._db)
self.viaf = VIAFClient(self._db)
@property
def input_data_source(self):
return DataSource.lookup(self._db, DataSource.OCLC_LINKED_DATA)
def do_run(self):
"""Re-runs OCLC Linked Data coverage provider to get viafs. Fetches
author information and recalculates presentation."""
identifiers = self.fetch_authorless_threem_identifiers()
self.delete_coverage_records(identifiers)
self.ensure_isbn_identifier(identifiers)
for identifier in identifiers:
self.coverage.ensure_coverage(identifier)
self.merge_contributors(identifier)
# Recalculate everything so the contributors can be seen.
for contributor in identifier.primary_edition.contributors:
self.viaf.process_contributor(contributor)
identifier.primary_edition.calculate_presentation()
if identifier.licensed_through:
identifier.licensed_through.calculate_work()
def fetch_authorless_threem_identifiers(self):
"""Returns a list of ThreeM identifiers that don't have contributors"""
qu = self._db.query(Identifier).join(Identifier.primarily_identifies)
qu = qu.outerjoin(Edition.contributions).filter(Contribution.id==None)
qu = qu.filter(Identifier.type == Identifier.THREEM_ID)
return qu.all()
def delete_coverage_records(self, identifiers):
"""Deletes existing OCLC Linked Data coverage records to re-run and
capture author data"""
t1 = self._db.begin_nested()
for identifier in identifiers:
for coverage_record in identifier.coverage_records:
if coverage_record.data_source == self.input_data_source:
self._db.delete(coverage_record)
t1.commit()
def ensure_isbn_identifier(self, identifiers):
"""Runs OCLCClassify to get ISBN numbers if they're not available."""
identifiers_without_isbn = []
for identifier in identifiers:
equivalencies = identifier.equivalencies
equivalent_types = [eq.output.type for eq in equivalencies]
if Identifier.ISBN not in equivalent_types:
identifiers_without_isbn.append(identifier)
for identifier in identifiers_without_isbn:
self.oclc_classify.ensure_coverage(identifier)
def merge_contributors(self, identifier):
"""Gives a ThreeM primary edition any contributors found via OCLC-LD"""
qu = self._db.query(Identifier).join(Identifier.inbound_equivalencies)
qu = qu.filter(or_(
Identifier.type == Identifier.OCLC_WORK,
Identifier.type == Identifier.OCLC_NUMBER
)).filter(Equivalency.input_id == identifier.id)
oclc_contributions = []
for oclc_identifier in qu.all():
editions = oclc_identifier.primarily_identifies
for edition in editions:
oclc_contributions += edition.contributions
for contribution in oclc_contributions:
for edition in identifier.primarily_identifies:
edition.add_contributor(contribution.contributor, contribution.role)
