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