def handle(self, **options):
badworks = Work.objects.exclude(language__regex=iso639)
badworks = badworks.exclude(language__regex=lang_and_locale)
self.stdout.write('{} works to fix'.format(badworks.count()))
for work in badworks:
language = lang_to_language_code(work.language)
work.language = language if language else 'xx'
work.save()
def get_language(self):
langlabel = self.doc.find(string='Language')
lang = langlabel.parent.parent.find_next_sibling() if langlabel else ''
lang = lang.get_text() if lang else ''
lang = lang_to_language_code(lang) if lang else ''
if lang:
self.set('language', lang)
else:
super(UbiquityScraper, self).get_language()
def load_from_pandata(self, metadata, work=None):
''' metadata is a Pandata object'''
#find an work to associate
edition = None
has_ed_id = False
if metadata.url:
new_ids = [('http', 'http', metadata.url)]
else:
new_ids = []
for (identifier, id_code) in IDTABLE:
# note that the work chosen is the last associated
value = metadata.edition_identifiers.get(identifier, None)
value = identifier_cleaner(id_code)(value)
if not value:
value = metadata.identifiers.get(identifier, None)
if value:
if id_code not in WORK_IDENTIFIERS:
has_ed_id = True
value = value[0] if isinstance(value, list) else value
try:
id = models.Identifier.objects.get(type=id_code, value=value)
if work and id.work and id.work_id is not work.id:
# dangerous! merge newer into older
if work.id < id.work_id:
work = merge_works(work, id.work)
else:
work = merge_works(id.work, work)
else:
work = id.work
if id.edition and not edition:
edition = id.edition
except models.Identifier.DoesNotExist:
if id_code != 'edid' or not has_ed_id: #last in loop
# only need to create edid if there is no edition id for the edition
new_ids.append((identifier, id_code, value))
if not work:
if metadata.title:
language = lang_to_language_code(metadata.language)
work = models.Work.objects.create(title=metadata.title, language=language if language else 'xx')
else:
return None
if not edition:
if metadata.edition_note:
(note, created) = models.EditionNote.objects.get_or_create(note=metadata.edition_note)
else:
note = None
edition = models.Edition.objects.create(
title=metadata.title,
work=work,
note=note,
)
for (identifier, id_code, value) in new_ids:
models.Identifier.set(
type=id_code,
value=value,
edition=edition if id_code not in WORK_IDENTIFIERS else None,
work=work,
)
if metadata.publisher: #always believe yaml
edition.set_publisher(metadata.publisher)
if metadata.publication_date: #always believe yaml
edition.publication_date = metadata.publication_date
#be careful about overwriting the work description
if metadata.description and len(metadata.description) > len(work.description):
# don't over-write reasonably long descriptions
if len(work.description) < 500:
work.description = metadata.description
if metadata.creator and not edition.authors.count():
edition.authors.clear()
for key in metadata.creator.keys():
creators = metadata.creator[key]
rel_code = inverse_marc_rels.get(key, None)
if not rel_code:
rel_code = inverse_marc_rels.get(key.rstrip('s'), 'auth')
creators = creators if isinstance(creators, list) else [creators]
for creator in creators:
edition.add_author(unreverse_name(creator.get('agent_name', '')), relation=rel_code)
for yaml_subject in metadata.subjects: #always add yaml subjects (don't clear)
if isinstance(yaml_subject, tuple):
(authority, heading) = yaml_subject
elif isinstance(yaml_subject, str) or isinstance(yaml_subject, unicode):
(authority, heading) = ('', yaml_subject)
else:
continue
subject = models.Subject.set_by_name(heading, work=work, authority=authority)
# the default edition uses the first cover in covers.
for cover in metadata.covers:
if cover.get('image_path', False):
edition.cover_image = urljoin(self.base_url, cover['image_path'])
break
elif cover.get('image_url', False):
edition.cover_image = cover['image_url']
break
work.save()
edition.save()
return edition
def doab_lang_to_iso_639_1(lang):
lang = lang_to_language_code(lang)
return lang if lang else 'xx'