apps/oai/handlers.py

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