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