src/oai/handlers.py

   1 # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later.
   2 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
   3 #
   4 from datetime import datetime
   5 from lxml import etree
   6
   7 from django.conf import settings
   8 from django.contrib.contenttypes.models import ContentType
   9 from django.contrib.sites.models import Site
  10 from django.utils import timezone
  11 from oaipmh import server, common, metadata, error
  12
  13 from api.handlers import WL_BASE
  14 from api.models import Deleted
  15 from catalogue.models import Book
  16 from librarian import WLURI
  17
  18
  19 def make_time_naive(d):
  20     return timezone.localtime(d).replace(tzinfo=None)
  21
  22 WL_DC_READER_XPATH = '(.|*)/rdf:RDF/rdf:Description/%s/text()'
  23 wl_dc_reader = metadata.MetadataReader(
  24     fields={
  25         'title':       ('textList', WL_DC_READER_XPATH % 'dc:title'),
  26         'creator':     ('textList', WL_DC_READER_XPATH % 'dc:creator'),
  27         'subject':     ('textList', (WL_DC_READER_XPATH + ' | ' + WL_DC_READER_XPATH + ' | ' + WL_DC_READER_XPATH) %
  28                         ('dc:subject.period', 'dc:subject.type', 'dc:subject.genre')),
  29         'description': ('textList', WL_DC_READER_XPATH % 'dc:description'),
  30         'publisher':   ('textList', WL_DC_READER_XPATH % 'dc:publisher'),
  31         'contributor': ('textList', (WL_DC_READER_XPATH + ' | ' + WL_DC_READER_XPATH + ' | ' + WL_DC_READER_XPATH) %
  32                         ('dc:contributor.editor', 'dc:contributor.translator', 'dc:contributor.technical_editor')),
  33         'date':        ('textList', WL_DC_READER_XPATH % 'dc:date'),
  34         'type':        ('textList', WL_DC_READER_XPATH % 'dc:type'),
  35         'format':      ('textList', WL_DC_READER_XPATH % 'dc:format'),
  36         'identifier':  ('textList', WL_DC_READER_XPATH % 'dc:identifier.url'),
  37         'source':      ('textList', WL_DC_READER_XPATH % 'dc:source'),
  38         'language':    ('textList', WL_DC_READER_XPATH % 'dc:language'),
  39         # 'isPartOf':    ('textList', 'rdf:RDF/rdf:Description/dc:relation.isPartOf/text()'),
  40         'hasPart':     ('textList', WL_DC_READER_XPATH % 'dc:relation.hasPart'),
  41         # 'relation':    ('textList', 'rdf:RDF/rdf:Description/dc:relation/text()'),
  42         # 'coverage':    ('textList', 'rdf:RDF/rdf:Description/dc:coverage/text()'),
  43         'rights':      ('textList', WL_DC_READER_XPATH % 'dc:rights')
  44     },
  45     namespaces={
  46         'dc': 'http://purl.org/dc/elements/1.1/',
  47         'rdf': 'http://www.w3.org/1999/02/22-rdf-syntax-ns#'}
  48     )
  49
  50
  51 NS_DCTERMS = "http://purl.org/dc/terms/"
  52
  53
  54 def nsdcterms(name):
  55     return '{%s}%s' % (NS_DCTERMS, name)
  56
  57
  58 class Catalogue(common.ResumptionOAIPMH):
  59     TAG_CATEGORIES = ['author', 'epoch', 'kind', 'genre']
  60
  61     def __init__(self, metadata_registry):
  62         super(Catalogue, self).__init__()
  63         self.metadata_registry = metadata_registry
  64
  65         self.oai_id = "oai:" + Site.objects.get_current().domain + ":%s"
  66
  67         # earliest change
  68         year_zero = timezone.make_aware(datetime(1990, 1, 1, 0, 0, 0), timezone.utc)
  69
  70         try:
  71             earliest_change = Book.objects.filter(findable=True, preview=False).order_by('changed_at')[0].changed_at
  72         except IndexError:
  73             earliest_change = year_zero
  74
  75         try:
  76             earliest_delete = \
  77                 Deleted.objects.exclude(slug__exact='').order_by('deleted_at')[0].deleted_at
  78         except IndexError:
  79             earliest_delete = year_zero
  80
  81         self.earliest_datestamp = earliest_change if earliest_change <= earliest_delete else earliest_delete
  82
  83     @staticmethod
  84     def metadata(book):
  85         try:
  86             xml = etree.parse(book.xml_file)
  87         finally:
  88             book.xml_file.close()
  89         md = wl_dc_reader(xml)
  90         m = md.getMap()
  91         if book.parent:
  92             m['isPartOf'] = [str(WLURI.from_slug(book.parent.slug))]
  93         return m
  94
  95     def record_for_book(self, book, headers_only=False):
  96         meta = None
  97         identifier = self.slug_to_identifier(book.slug)
  98         if isinstance(book, Book):
  99             # setSpec = map(self.tag_to_setspec, book.tags.filter(category__in=self.TAG_CATEGORIES))
 100             header = common.Header(None, identifier, make_time_naive(book.changed_at), [], False)
 101             if not headers_only:
 102                 meta = common.Metadata(None, self.metadata(book))
 103             about = None
 104         elif isinstance(book, Deleted):
 105             header = common.Header(None, identifier, make_time_naive(book.deleted_at), [], True)
 106             if not headers_only:
 107                 meta = common.Metadata(None, {})
 108             about = None
 109         else:
 110             raise TypeError('Unknown book class')
 111         if headers_only:
 112             return header
 113         return header, meta, about
 114
 115     def identify(self, **kw):
 116         ident = common.Identify(
 117             'Wolne Lektury',  # generate
 118             '%s/oaipmh' % str(WL_BASE),  # generate
 119             '2.0',  # version
 120             [m[1] for m in settings.MANAGERS],  # adminEmails
 121             make_time_naive(self.earliest_datestamp),  # earliest datestamp of any change
 122             'persistent',  # deletedRecord
 123             'YYYY-MM-DDThh:mm:ssZ',  # granularity
 124             ['identity'],  # compression
 125             []  # descriptions
 126             )
 127         return ident
 128
 129     def books(self, tag, from_, until):
 130         if tag:
 131             # we do not support sets, since they are problematic for deleted books.
 132             raise error.NoSetHierarchyError("Wolne Lektury does not support sets.")
 133             # books = Book.tagged.with_all([tag])
 134         else:
 135             books = Book.objects.filter(findable=True, preview=False)
 136         deleted = Deleted.objects.exclude(slug__exact='')
 137
 138         books = books.order_by('changed_at')
 139         deleted = deleted.order_by('deleted_at')
 140         if from_:
 141             books = books.filter(changed_at__gte=from_)
 142             deleted = deleted.filter(deleted_at__gte=from_)
 143         if until:
 144             books = books.filter(changed_at__lte=until)
 145             deleted = deleted.filter(deleted_at__lte=until)
 146         return list(books) + list(deleted)
 147
 148     # @staticmethod
 149     # def tag_to_setspec(tag):
 150     #     return "%s:%s" % (tag.category, tag.slug)
 151
 152     # @staticmethod
 153     # def setspec_to_tag(s):
 154     #     if not s: return None
 155     #     cs = s.split(':')
 156     #     if len(cs) == 2:
 157     #         if not cs[0] in Catalogue.TAG_CATEGORIES:
 158     #             raise error.NoSetHierarchyError("No category part in set")
 159     #         tag = Tag.objects.get(slug=cs[1], category=cs[0])
 160     #         return tag
 161     #     raise error.NoSetHierarchyError("Setspec should have two components: category:slug")
 162
 163     def getRecord(self, **kw):
 164         """Returns (header, metadata, about) for given record."""
 165         slug = self.identifier_to_slug(kw['identifier'])
 166         try:
 167             book = Book.objects.get(slug=slug)
 168             return self.record_for_book(book)
 169         except Book.DoesNotExist:
 170             book_type = ContentType.objects.get_for_model(Book)
 171             try:
 172                 deleted_book = Deleted.objects.get(content_type=book_type, slug=slug)
 173             except:
 174                 raise error.IdDoesNotExistError("No item for this identifier")
 175             return self.record_for_book(deleted_book)
 176
 177     def validate_kw(self, kw):
 178         if 'resumptionToken' in kw:
 179             raise error.BadResumptionTokenError("No resumption token support at this point")
 180         if 'metadataPrefix' in kw and not self.metadata_registry.hasWriter(kw['metadataPrefix']):
 181             raise error.CannotDisseminateFormatError("This format is not supported")
 182
 183     def identifier_to_slug(self, ident):
 184         return ident.split(':')[-1]
 185
 186     def slug_to_identifier(self, slug):
 187         return self.oai_id % slug
 188
 189     def listIdentifiers(self, **kw):
 190         self.validate_kw(kw)
 191         records = [self.record_for_book(book, headers_only=True) for
 192                    book in self.books(None, kw.get('from_'), kw.get('until'))]
 193         return records, None
 194
 195     def listRecords(self, **kw):
 196         """
 197             can get a resumptionToken kw.
 198             returns result, token
 199         """
 200         self.validate_kw(kw)
 201         records = [self.record_for_book(book) for
 202                    book in self.books(None,
 203                            kw.get('from_', None),
 204                            kw.get('until', None))]
 205
 206         return records, None
 207
 208     def listMetadataFormats(self, **kw):
 209         formats = [
 210             ('oai_dc',
 211              'http://www.openarchives.org/OAI/2.0/oai_dc.xsd',
 212              server.NS_OAIDC),
 213             ('qdc',
 214              'http://dublincore.org/schemas/xmls/qdc/2006/01/06/dcterms.xsd',
 215              NS_DCTERMS)]
 216         if 'identifier' in kw:
 217             slug = self.identifier_to_slug(kw['identifier'])
 218             try:
 219                 Book.objects.get(slug=slug)
 220                 return formats
 221             except Book.DoesNotExist:
 222                 try:
 223                     Deleted.objects.get(slug=slug)
 224                     return []
 225                 except Deleted.DoesNotExist:
 226                     raise error.IdDoesNotExistError("This id does not exist")
 227         else:
 228             return formats
 229
 230     def listSets(self, **kw):
 231         raise error.NoSetHierarchyError("Wolne Lektury does not support sets.")
 232         # tags = []
 233         # for category in Catalogue.TAG_CATEGORIES:
 234         #     for tag in Tag.objects.filter(category=category):
 235         #         tags.append(("%s:%s" % (tag.category, tag.slug),
 236         #                      tag.name,
 237         #                      tag.description))
 238         # return tags, None