def categories(cls, work):
"""Send out _all_ categories for the work.
(So long as the category type has a URI associated with it in
Subject.uri_lookup.)
"""
_db = Session.object_session(work)
by_scheme_and_term = dict()
identifier_ids = work.all_identifier_ids()
classifications = Identifier.classifications_for_identifier_ids(
_db, identifier_ids)
for c in classifications:
subject = c.subject
if subject.type in Subject.uri_lookup:
scheme = Subject.uri_lookup[subject.type]
term = subject.identifier
weight_field = AtomFeed.schema_("ratingValue")
key = (scheme, term)
if not key in by_scheme_and_term:
value = dict(term=subject.identifier)
if subject.name:
value['label'] = subject.name
value[weight_field] = 0
by_scheme_and_term[key] = value
by_scheme_and_term[key][weight_field] += c.weight
# Collapse by_scheme_and_term to by_scheme
by_scheme = defaultdict(list)
for (scheme, term), value in by_scheme_and_term.items():
by_scheme[scheme].append(value)
by_scheme.update(super(VerboseAnnotator, cls).categories(work))
return by_scheme
def explain_identifier(cls, identifier, primary, seen, strength, level):
indent = " " * level
if primary:
ident = "Primary identifier"
else:
ident = "Identifier"
if primary:
strength = 1
output = "%s %s: %s/%s (q=%s)" % (indent, ident, identifier.type, identifier.identifier, strength)
print output.encode("utf8")
_db = Session.object_session(identifier)
classifications = Identifier.classifications_for_identifier_ids(
_db, [identifier.id])
for classification in classifications:
subject = classification.subject
genre = subject.genre
if genre:
genre = genre.name
else:
genre = "(!genre)"
#print "%s %s says: %s/%s %s w=%s" % (
# indent, classification.data_source.name,
# subject.identifier, subject.name, genre, classification.weight
#)
seen.add(identifier)
for equivalency in identifier.equivalencies:
if equivalency.id in seen:
continue
seen.add(equivalency.id)
output = equivalency.output
cls.explain_identifier(output, False, seen,
equivalency.strength, level+1)
