X-Git-Url: https://git.mdrn.pl/wolnelektury.git/blobdiff_plain/681be17b1c53f4b15838b3c4a53b0b01ee0a86e5..0bd2f26c7d4a349bf9ae9377a945baf39d28af00:/apps/opds/views.py diff --git a/apps/opds/views.py b/apps/opds/views.py index 9431684b3..c529e0272 100644 --- a/apps/opds/views.py +++ b/apps/opds/views.py @@ -2,7 +2,6 @@ # This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later. # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. # -from base64 import b64encode import os.path from urlparse import urljoin @@ -13,11 +12,19 @@ from django.utils.feedgenerator import Atom1Feed from django.conf import settings from django.http import Http404 from django.contrib.sites.models import Site +from django.utils.functional import lazy from basicauth import logged_in_or_basicauth, factory_decorator from catalogue.models import Book, Tag -from catalogue.views import books_starting_with +from search.views import Search +import operator +import logging +import re + +log = logging.getLogger('opds') + +from stats.utils import piwik_track _root_feeds = ( { @@ -58,21 +65,22 @@ _root_feeds = ( ) +current_domain = lazy(lambda: Site.objects.get_current().domain, str)() def full_url(url): - return urljoin("http://%s" % Site.objects.get_current().domain, url) + return urljoin("http://%s" % current_domain, url) class OPDSFeed(Atom1Feed): link_rel = u"subsection" link_type = u"application/atom+xml" - _book_parent_img = full_url(os.path.join(settings.STATIC_URL, "img/book-parent.png")) + _book_parent_img = lazy(lambda: full_url(os.path.join(settings.STATIC_URL, "img/book-parent.png")), str)() try: _book_parent_img_size = unicode(os.path.getsize(os.path.join(settings.STATIC_ROOT, "img/book-parent.png"))) except: _book_parent_img_size = '' - _book_img = full_url(os.path.join(settings.STATIC_URL, "img/book.png")) + _book_img = lazy(lambda: full_url(os.path.join(settings.STATIC_URL, "img/book.png")), str)() try: _book_img_size = unicode(os.path.getsize(os.path.join(settings.STATIC_ROOT, "img/book.png"))) except: @@ -85,7 +93,7 @@ class OPDSFeed(Atom1Feed): {u"href": reverse("opds_authors"), u"rel": u"start", u"type": u"application/atom+xml"}) - handler.addQuickElement(u"link", None, + handler.addQuickElement(u"link", None, {u"href": full_url(os.path.join(settings.STATIC_URL, "opensearch.xml")), u"rel": u"search", u"type": u"application/opensearchdescription+xml"}) @@ -105,6 +113,7 @@ class OPDSFeed(Atom1Feed): u"length": self._book_parent_img_size, u"type": u"image/png"}) if item['pubdate'] is not None: + # FIXME: rfc3339_date is undefined, is this ever run? handler.addQuickElement(u"updated", rfc3339_date(item['pubdate']).decode('utf-8')) # Author information. @@ -121,6 +130,7 @@ class OPDSFeed(Atom1Feed): if item['unique_id'] is not None: unique_id = item['unique_id'] else: + # FIXME: get_tag_uri is undefined, is this ever run? unique_id = get_tag_uri(item['link'], item['pubdate']) handler.addQuickElement(u"id", unique_id) @@ -181,19 +191,19 @@ class AcquisitionFeed(Feed): return u'' def item_enclosure_url(self, book): - return full_url(book.root_ancestor.epub_file.url) + return full_url(book.epub_file.url) if book.epub_file else None def item_enclosure_length(self, book): - return book.root_ancestor.epub_file.size - + return book.epub_file.size if book.epub_file else None +@piwik_track class RootFeed(Feed): feed_type = OPDSFeed title = u'Wolne Lektury' - link = u'http://www.wolnelektury.pl/' + link = u'http://wolnelektury.pl/' description = u"Spis utworów na stronie http://WolneLektury.pl" author_name = u"Wolne Lektury" - author_link = u"http://www.wolnelektury.pl/" + author_link = u"http://wolnelektury.pl/" def items(self): return _root_feeds @@ -207,16 +217,16 @@ class RootFeed(Feed): def item_description(self, item): return item['description'] - +@piwik_track class ByCategoryFeed(Feed): feed_type = OPDSFeed - link = u'http://www.wolnelektury.pl/' + link = u'http://wolnelektury.pl/' description = u"Spis utworów na stronie http://WolneLektury.pl" author_name = u"Wolne Lektury" - author_link = u"http://www.wolnelektury.pl/" + author_link = u"http://wolnelektury.pl/" def get_object(self, request, category): - feed = [feed for feed in _root_feeds if feed['category']==category] + feed = [feed for feed in _root_feeds if feed['category'] == category] if feed: feed = feed[0] else: @@ -228,7 +238,7 @@ class ByCategoryFeed(Feed): return feed['title'] def items(self, feed): - return (tag for tag in Tag.objects.filter(category=feed['category']) if tag.get_count() > 0) + return Tag.objects.filter(category=feed['category']).exclude(items=None) def item_title(self, item): return item.name @@ -239,7 +249,7 @@ class ByCategoryFeed(Feed): def item_description(self): return u'' - +@piwik_track class ByTagFeed(AcquisitionFeed): def link(self, tag): return tag.get_absolute_url() @@ -254,22 +264,17 @@ class ByTagFeed(AcquisitionFeed): return get_object_or_404(Tag, category=category, slug=slug) def items(self, tag): - books = Book.tagged.with_any([tag]) - l_tags = Tag.objects.filter(category='book', slug__in=[book.book_tag_slug() for book in books]) - descendants_keys = [book.pk for book in Book.tagged.with_any(l_tags)] - if descendants_keys: - books = books.exclude(pk__in=descendants_keys) - - return books + return Book.tagged_top_level([tag]) -#@factory_decorator(logged_in_or_basicauth()) +@factory_decorator(logged_in_or_basicauth()) +@piwik_track class UserFeed(Feed): feed_type = OPDSFeed link = u'http://www.wolnelektury.pl/' description = u"Półki użytkownika na stronie http://WolneLektury.pl" author_name = u"Wolne Lektury" - author_link = u"http://www.wolnelektury.pl/" + author_link = u"http://wolnelektury.pl/" def get_object(self, request): return request.user @@ -278,7 +283,7 @@ class UserFeed(Feed): return u"Półki użytkownika %s" % user.username def items(self, user): - return (tag for tag in Tag.objects.filter(category='set', user=user) if tag.get_count() > 0) + return Tag.objects.filter(category='set', user=user).exclude(items=None) def item_title(self, item): return item.name @@ -289,11 +294,9 @@ class UserFeed(Feed): def item_description(self): return u'' -# no class decorators in python 2.5 -UserFeed = factory_decorator(logged_in_or_basicauth())(UserFeed) - -#@factory_decorator(logged_in_or_basicauth()) +@factory_decorator(logged_in_or_basicauth()) +@piwik_track class UserSetFeed(AcquisitionFeed): def link(self, tag): return tag.get_absolute_url() @@ -310,23 +313,124 @@ class UserSetFeed(AcquisitionFeed): def items(self, tag): return Book.tagged.with_any([tag]) -# no class decorators in python 2.5 -UserSetFeed = factory_decorator(logged_in_or_basicauth())(UserSetFeed) - +@piwik_track class SearchFeed(AcquisitionFeed): description = u"Wyniki wyszukiwania na stronie WolneLektury.pl" title = u"Wyniki wyszukiwania" - + + QUOTE_OR_NOT = r'(?:(?=["])"([^"]+)"|([^ ]+))' + INLINE_QUERY_RE = re.compile( + r"author:" + QUOTE_OR_NOT + + "|translator:" + QUOTE_OR_NOT + + "|title:" + QUOTE_OR_NOT + + "|categories:" + QUOTE_OR_NOT + + "|description:" + QUOTE_OR_NOT + + "|text:" + QUOTE_OR_NOT + ) + MATCHES = { + 'author': (0, 1), + 'translator': (2, 3), + 'title': (4, 5), + 'categories': (6, 7), + 'description': (8, 9), + 'text': (10, 11), + } + + PARAMS_TO_FIELDS = { + 'author': 'authors', + 'translator': 'translators', + # 'title': 'title', + 'categories': 'tag_name_pl', + 'description': 'text', + # 'text': 'text', + } + + ATOM_PLACEHOLDER = re.compile(r"^{(atom|opds):\w+}$") + def get_object(self, request): - return request.GET.get('q', '') + """ + For OPDS 1.1 We should handle a query for search terms + and criteria provided either as opensearch or 'inline' query. + OpenSearch defines fields: atom:author, atom:contributor (treated as translator), + atom:title. Inline query provides author, title, categories (treated as book tags), + description (treated as content search terms). + + if search terms are provided, we shall search for books + according to Hint information (from author & contributror & title). + + but if search terms are empty, we should do a different search + (perhaps for is_book=True) + + """ + + query = request.GET.get('q', '') + + inline_criteria = re.findall(self.INLINE_QUERY_RE, query) + if inline_criteria: + remains = re.sub(self.INLINE_QUERY_RE, '', query) + remains = re.sub(r'[ \t]+', ' ', remains) + + def get_criteria(criteria, name): + for c in criteria: + for p in self.MATCHES[name]: + if c[p]: + if p % 2 == 0: + return c[p].replace('+', ' ') + return c[p] + return None + + criteria = dict(map( + lambda cn: (cn, get_criteria(inline_criteria, cn)), + ['author', 'translator', 'title', 'categories', + 'description', 'text'])) + query = remains + # empty query and text set case? + log.debug("Inline query = [%s], criteria: %s" % (query, criteria)) + else: + def remove_dump_data(val): + """Some clients don't get opds placeholders and just send them.""" + if self.ATOM_PLACEHOLDER.match(val): + return '' + return val + + criteria = dict([(cn, remove_dump_data(request.GET.get(cn, ''))) + for cn in self.MATCHES.keys()]) + # query is set above. + log.debug("Inline query = [%s], criteria: %s" % (query, criteria)) + + srch = Search() + + book_hit_filter = srch.index.Q(book_id__any=True) + filters = [book_hit_filter] + [srch.index.Q( + **{self.PARAMS_TO_FIELDS.get(cn, cn): criteria[cn]} + ) for cn in self.MATCHES.keys() if cn in criteria + if criteria[cn]] + + if query: + q = srch.index.query( + reduce(operator.or_, + [srch.index.Q(**{self.PARAMS_TO_FIELDS.get(cn, cn): query}) + for cn in self.MATCHES.keys()], + srch.index.Q())) + else: + q = srch.index.query(srch.index.Q()) + + q = srch.apply_filters(q, filters).field_limit(score=True, fields=['book_id']) + results = q.execute() + + book_scores = dict([(r['book_id'], r['score']) for r in results]) + books = Book.objects.filter(id__in=set([r['book_id'] for r in results])) + books = list(books) + books.sort(reverse=True, key=lambda book: book_scores[book.id]) + return books def get_link(self, query): - return "%s?q=%s" % (reverse('search'), query) + return "%s?q=%s" % (reverse('search'), query) - def items(self, query): + def items(self, books): try: - return books_starting_with(query) + return books except ValueError: # too short a query return []