X-Git-Url: https://git.mdrn.pl/wolnelektury.git/blobdiff_plain/d467f1fa349730919e9fe906287e63563c888063..67df302167539474edd02b57c2e8bd30349d7625:/apps/catalogue/views.py?ds=inline diff --git a/apps/catalogue/views.py b/apps/catalogue/views.py index de72b046a..556bae1fe 100644 --- a/apps/catalogue/views.py +++ b/apps/catalogue/views.py @@ -8,6 +8,8 @@ import sys import pprint import traceback import re +import itertools +from operator import itemgetter from django.conf import settings from django.template import RequestContext @@ -26,6 +28,7 @@ from django.utils.encoding import force_unicode from django.utils.http import urlquote_plus from django.views.decorators import cache from django.utils.translation import ugettext as _ +from django.views.generic.list_detail import object_list from catalogue import models from catalogue import forms @@ -42,119 +45,172 @@ class LazyEncoder(simplejson.JSONEncoder): return force_unicode(obj) return obj +# shortcut for JSON reponses +class JSONResponse(HttpResponse): + def __init__(self, data={}, callback=None, **kwargs): + # get rid of mimetype + kwargs.pop('mimetype', None) + data = simplejson.dumps(data) + if callback: + data = callback + "(" + data + ");" + super(JSONResponse, self).__init__(data, mimetype="application/json", **kwargs) -def main_page(request): + +def main_page(request): if request.user.is_authenticated(): shelves = models.Tag.objects.filter(category='set', user=request.user) new_set_form = forms.NewSetForm() - extra_where = "NOT catalogue_tag.category = 'set'" - tags = models.Tag.objects.usage_for_model(models.Book, counts=True, extra={'where': [extra_where]}) - fragment_tags = models.Tag.objects.usage_for_model(models.Fragment, counts=True, - extra={'where': ["catalogue_tag.category = 'theme'"] + [extra_where]}) + + tags = models.Tag.objects.exclude(category__in=('set', 'book')) + for tag in tags: + tag.count = tag.get_count() categories = split_tags(tags) - + fragment_tags = categories.get('theme', []) + form = forms.SearchForm() return render_to_response('catalogue/main_page.html', locals(), context_instance=RequestContext(request)) def book_list(request): - books = models.Book.objects.all() form = forms.SearchForm() - - books_by_first_letter = SortedDict() - for book in books: - books_by_first_letter.setdefault(book.title[0], []).append(book) - + + books_by_parent = {} + for book in models.Book.objects.all().order_by('parent_number'): + books_by_parent.setdefault(book.parent, []).append(book) + + orphans = [] + books_by_author = SortedDict() + for tag in models.Tag.objects.filter(category='author'): + books_by_author[tag] = [] + + for book in books_by_parent[None]: + authors = list(book.tags.filter(category='author')) + if authors: + for author in authors: + books_by_author[author].append(book) + else: + orphans.append(book) + return render_to_response('catalogue/book_list.html', locals(), context_instance=RequestContext(request)) +def differentiate_tags(request, tags, ambiguous_slugs): + beginning = '/'.join(tag.url_chunk for tag in tags) + unparsed = '/'.join(ambiguous_slugs[1:]) + options = [] + for tag in models.Tag.objects.exclude(category='book').filter(slug=ambiguous_slugs[0]): + options.append({ + 'url_args': '/'.join((beginning, tag.url_chunk, unparsed)).strip('/'), + 'tags': [tag] + }) + return render_to_response('catalogue/differentiate_tags.html', + {'tags': tags, 'options': options, 'unparsed': ambiguous_slugs[1:]}, + context_instance=RequestContext(request)) + + def tagged_object_list(request, tags=''): - # Prevent DoS attacks on our database - if len(tags.split('/')) > 6: - raise Http404 - try: tags = models.Tag.get_tag_list(tags) except models.Tag.DoesNotExist: raise Http404 - + except models.Tag.MultipleObjectsReturned, e: + return differentiate_tags(request, e.tags, e.ambiguous_slugs) + + try: + if len(tags) > settings.MAX_TAG_LIST: + raise Http404 + except AttributeError: + pass + if len([tag for tag in tags if tag.category == 'book']): raise Http404 - + theme_is_set = [tag for tag in tags if tag.category == 'theme'] - shelf_is_set = len(tags) == 1 and tags[0].category == 'set' - my_shelf_is_set = shelf_is_set and request.user.is_authenticated() and request.user == shelf[0].user - - objects = only_author = pd_counter = categories = None - + shelf_is_set = [tag for tag in tags if tag.category == 'set'] + only_shelf = shelf_is_set and len(tags) == 1 + only_my_shelf = only_shelf and request.user.is_authenticated() and request.user == tags[0].user + + objects = only_author = pd_counter = None + categories = {} + if theme_is_set: shelf_tags = [tag for tag in tags if tag.category == 'set'] fragment_tags = [tag for tag in tags if tag.category != 'set'] fragments = models.Fragment.tagged.with_all(fragment_tags) - + if shelf_tags: - books = models.Book.tagged.with_all(shelf_tags) - l_tags = [models.Tag.objects.get(slug = 'l-' + book.slug) for book in books] - fragments = (fragment for fragment in models.Fragment.tagged.with_any(l_tags) if fragment in fragments) - - fragment_keys = (fragment.pk for fragment in fragments) + books = models.Book.tagged.with_all(shelf_tags).order_by() + l_tags = models.Tag.objects.filter(category='book', slug__in=[book.book_tag_slug() for book in books]) + fragments = models.Fragment.tagged.with_any(l_tags, fragments) + + # newtagging goes crazy if we just try: + #related_tags = models.Tag.objects.usage_for_queryset(fragments, counts=True, + # extra={'where': ["catalogue_tag.category != 'book'"]}) + fragment_keys = [fragment.pk for fragment in fragments] if fragment_keys: related_tags = models.Fragment.tags.usage(counts=True, - filters={'pk__in': fragment_keys}, + filters={'pk__in': fragment_keys}, extra={'where': ["catalogue_tag.category != 'book'"]}) related_tags = (tag for tag in related_tags if tag not in fragment_tags) categories = split_tags(related_tags) - + objects = fragments else: - books = models.Book.tagged.with_all(tags) - l_tags = [models.Tag.objects.get(slug = 'l-' + book.slug) for book in books] - book_keys = [book.pk for book in books] - if book_keys: - related_tags = models.Book.tags.usage(counts=True, - filters={'pk__in': book_keys}, - extra={'where': ["catalogue_tag.category NOT IN ('set', 'book', 'theme')"]}) - categories = split_tags(related_tags) - - fragment_keys = [fragment.pk for fragment in models.Fragment.tagged.with_any(l_tags)] - if fragment_keys: - categories['theme'] = models.Fragment.tags.usage(counts=True, - filters={'pk__in': fragment_keys}, - extra={'where': ["catalogue_tag.category = 'theme'"]}) - - books = books.exclude(parent__in = book_keys) - objects = books - + # get relevant books and their tags + objects = models.Book.tagged.with_all(tags).order_by() + if not shelf_is_set: + # eliminate descendants + l_tags = models.Tag.objects.filter(category='book', slug__in=[book.book_tag_slug() for book in objects]) + descendants_keys = [book.pk for book in models.Book.tagged.with_any(l_tags)] + if descendants_keys: + objects = objects.exclude(pk__in=descendants_keys) + + # get related tags from `tag_counter` and `theme_counter` + related_counts = {} + tags_pks = [tag.pk for tag in tags] + for book in objects: + for tag_pk, value in itertools.chain(book.tag_counter.iteritems(), book.theme_counter.iteritems()): + if tag_pk in tags_pks: + continue + related_counts[tag_pk] = related_counts.get(tag_pk, 0) + value + related_tags = models.Tag.objects.filter(pk__in=related_counts.keys()) + related_tags = [tag for tag in related_tags if tag not in tags] + for tag in related_tags: + tag.count = related_counts[tag.pk] + + categories = split_tags(related_tags) + del related_tags + if not objects: only_author = len(tags) == 1 and tags[0].category == 'author' pd_counter = only_author and tags[0].goes_to_pd() objects = models.Book.objects.none() - - return newtagging_views.tagged_object_list( + + return object_list( request, - tag_model=models.Tag, - queryset_or_model=objects, - tags=tags, + objects, template_name='catalogue/tagged_object_list.html', - extra_context = { + extra_context={ 'categories': categories, - 'shelf_is_set': shelf_is_set, + 'only_shelf': only_shelf, 'only_author': only_author, 'pd_counter': pd_counter, - 'user_is_owner': my_shelf_is_set, + 'only_my_shelf': only_my_shelf, 'formats_form': forms.DownloadFormatsForm(), - }, + + 'tags': tags, + } ) def book_fragments(request, book_slug, theme_slug): book = get_object_or_404(models.Book, slug=book_slug) - book_tag = get_object_or_404(models.Tag, slug='l-' + book_slug) - theme = get_object_or_404(models.Tag, slug=theme_slug) + book_tag = get_object_or_404(models.Tag, slug='l-' + book_slug, category='book') + theme = get_object_or_404(models.Tag, slug=theme_slug, category='theme') fragments = models.Fragment.tagged.with_all([book_tag, theme]) - + form = forms.SearchForm() return render_to_response('catalogue/book_fragments.html', locals(), context_instance=RequestContext(request)) @@ -166,14 +222,25 @@ def book_detail(request, slug): except models.Book.DoesNotExist: return book_stub_detail(request, slug) - book_tag = get_object_or_404(models.Tag, slug = 'l-' + slug) + book_tag = book.book_tag() tags = list(book.tags.filter(~Q(category='set'))) categories = split_tags(tags) book_children = book.children.all().order_by('parent_number') - extra_where = "catalogue_tag.category = 'theme'" - book_themes = models.Tag.objects.related_for_model(book_tag, models.Fragment, counts=True, extra={'where': [extra_where]}) - extra_info = book.get_extra_info_value() + _book = book + parents = [] + while _book.parent: + parents.append(_book.parent) + _book = _book.parent + parents = reversed(parents) + + theme_counter = book.theme_counter + book_themes = models.Tag.objects.filter(pk__in=theme_counter.keys()) + for tag in book_themes: + tag.count = theme_counter[tag.pk] + + extra_info = book.get_extra_info_value() + form = forms.SearchForm() return render_to_response('catalogue/book_detail.html', locals(), context_instance=RequestContext(request)) @@ -183,10 +250,10 @@ def book_stub_detail(request, slug): book = get_object_or_404(models.BookStub, slug=slug) pd_counter = book.pd form = forms.SearchForm() - + return render_to_response('catalogue/book_stub_detail.html', locals(), context_instance=RequestContext(request)) - + def book_text(request, slug): book = get_object_or_404(models.Book, slug=slug) @@ -194,7 +261,7 @@ def book_text(request, slug): for fragment in book.fragments.all(): for theme in fragment.tags.filter(category='theme'): book_themes.setdefault(theme, []).append(fragment) - + book_themes = book_themes.items() book_themes.sort(key=lambda s: s[0].sort_key) return render_to_response('catalogue/book_text.html', locals(), @@ -207,64 +274,74 @@ def book_text(request, slug): def _no_diacritics_regexp(query): """ returns a regexp for searching for a query without diacritics - + should be locale-aware """ - names = {'a':u'ą', 'c':u'ć', 'e':u'ę', 'l': u'ł', 'n':u'ń', 'o':u'ó', 's':u'ś', 'z':u'ź|ż'} + names = { + u'a':u'aąĄ', u'c':u'cćĆ', u'e':u'eęĘ', u'l': u'lłŁ', u'n':u'nńŃ', u'o':u'oóÓ', u's':u'sśŚ', u'z':u'zźżŹŻ', + u'ą':u'ąĄ', u'ć':u'ćĆ', u'ę':u'ęĘ', u'ł': u'łŁ', u'ń':u'ńŃ', u'ó':u'óÓ', u'ś':u'śŚ', u'ź':u'źŹ', u'ż':u'żŻ' + } def repl(m): l = m.group() - return "(%s|%s)" % (l, names[l]) - return re.sub('[%s]'%(''.join(names.keys())), repl, query) + return u"(%s)" % '|'.join(names[l]) + return re.sub(u'[%s]' % (u''.join(names.keys())), repl, query) + +def unicode_re_escape(query): + """ Unicode-friendly version of re.escape """ + return re.sub('(?u)(\W)', r'\\\1', query) def _word_starts_with(name, prefix): """returns a Q object getting models having `name` contain a word starting with `prefix` + + We define word characters as alphanumeric and underscore, like in JS. + + Works for MySQL, PostgreSQL, Oracle. + For SQLite, _sqlite* version is substituted for this. """ kwargs = {} - if settings.DATABASE_ENGINE in ('mysql', 'postgresql_psycopg2', 'postgresql'): - prefix = _no_diacritics_regexp(re.escape(prefix)) - # we could use a [[:<:]] (word start), - # but we want both `xy` and `(xy` to catch `(xyz)` - kwargs['%s__iregex' % name] = u"(^|[^[:alpha:]])%s" % prefix - else: - # don't know how to do a generic regex - # checking for simple icontain instead - kwargs['%s__icontains' % name] = prefix + + prefix = _no_diacritics_regexp(unicode_re_escape(prefix)) + # can't use [[:<:]] (word start), + # but we want both `xy` and `(xy` to catch `(xyz)` + kwargs['%s__iregex' % name] = u"(^|[^[:alnum:]_])%s" % prefix + return Q(**kwargs) -def _tags_exact_matches(prefix, user): - book_stubs = models.BookStub.objects.filter(title__iexact = prefix) - books = models.Book.objects.filter(title__iexact = prefix) - book_stubs = filter(lambda x: x not in books, book_stubs) - tags = models.Tag.objects.filter(name__iexact = prefix) - if user.is_authenticated(): - tags = tags.filter(~Q(category='book') & (~Q(category='set') | Q(user=user))) - else: - tags = tags.filter(~Q(category='book') & ~Q(category='set')) +def _sqlite_word_starts_with(name, prefix): + """ version of _word_starts_with for SQLite - return list(books) + list(tags) + list(book_stubs) + SQLite in Django uses Python re module + """ + kwargs = {} + prefix = _no_diacritics_regexp(unicode_re_escape(prefix)) + kwargs['%s__iregex' % name] = ur"(^|(?<=[^\wąćęłńóśźżĄĆĘŁŃÓŚŹŻ]))%s" % prefix + return Q(**kwargs) -def _tags_starting_with(prefix, user): +if settings.DATABASE_ENGINE == 'sqlite3': + _word_starts_with = _sqlite_word_starts_with + + +def _tags_starting_with(prefix, user=None): + prefix = prefix.lower() book_stubs = models.BookStub.objects.filter(_word_starts_with('title', prefix)) books = models.Book.objects.filter(_word_starts_with('title', prefix)) book_stubs = filter(lambda x: x not in books, book_stubs) tags = models.Tag.objects.filter(_word_starts_with('name', prefix)) - if user.is_authenticated(): + if user and user.is_authenticated(): tags = tags.filter(~Q(category='book') & (~Q(category='set') | Q(user=user))) else: tags = tags.filter(~Q(category='book') & ~Q(category='set')) - return list(books) + list(tags) + list(book_stubs) - def _get_result_link(match, tag_list): if isinstance(match, models.Book) or isinstance(match, models.BookStub): return match.get_absolute_url() else: - return reverse('catalogue.views.tagged_object_list', - kwargs={'tags': '/'.join(tag.slug for tag in tag_list + [match])} + return reverse('catalogue.views.tagged_object_list', + kwargs={'tags': '/'.join(tag.url_chunk for tag in tag_list + [match])} ) def _get_result_type(match): @@ -272,38 +349,54 @@ def _get_result_type(match): type = 'book' else: type = match.category - return dict(models.TAG_CATEGORIES)[type] - + return type + + + +def find_best_matches(query, user=None): + """ Finds a Book, Tag or Bookstub best matching a query. + + Returns a with: + - zero elements when nothing is found, + - one element when a best result is found, + - more then one element on multiple exact matches + + Raises a ValueError on too short a query. + """ + + query = query.lower() + if len(query) < 2: + raise ValueError("query must have at least two characters") + + result = tuple(_tags_starting_with(query, user)) + exact_matches = tuple(res for res in result if res.name.lower() == query) + if exact_matches: + return exact_matches + else: + return result[:1] def search(request): tags = request.GET.get('tags', '') prefix = request.GET.get('q', '') - + try: tag_list = models.Tag.get_tag_list(tags) except: tag_list = [] - # Prefix must have at least 2 characters - if len(prefix) < 2: + try: + result = find_best_matches(prefix, request.user) + except ValueError: return render_to_response('catalogue/search_too_short.html', {'tags':tag_list, 'prefix':prefix}, context_instance=RequestContext(request)) - - result = _tags_exact_matches(prefix, request.user) - - if len(result) > 1: - # multiple exact matches - return render_to_response('catalogue/search_multiple_hits.html', + + if len(result) == 1: + return HttpResponseRedirect(_get_result_link(result[0], tag_list)) + elif len(result) > 1: + return render_to_response('catalogue/search_multiple_hits.html', {'tags':tag_list, 'prefix':prefix, 'results':((x, _get_result_link(x, tag_list), _get_result_type(x)) for x in result)}, context_instance=RequestContext(request)) - - if not result: - # no exact matches - result = _tags_starting_with(prefix, request.user) - - if result: - return HttpResponseRedirect(_get_result_link(result[0], tag_list)) else: return render_to_response('catalogue/search_no_hits.html', {'tags':tag_list, 'prefix':prefix}, context_instance=RequestContext(request)) @@ -314,9 +407,29 @@ def tags_starting_with(request): # Prefix must have at least 2 characters if len(prefix) < 2: return HttpResponse('') - - return HttpResponse('\n'.join(tag.name for tag in _tags_starting_with(prefix, request.user))) - + tags_list = [] + result = "" + for tag in _tags_starting_with(prefix, request.user): + if not tag.name in tags_list: + result += "\n" + tag.name + tags_list.append(tag.name) + return HttpResponse(result) + +def json_tags_starting_with(request, callback=None): + # Callback for JSONP + prefix = request.GET.get('q', '') + callback = request.GET.get('callback', '') + # Prefix must have at least 2 characters + if len(prefix) < 2: + return HttpResponse('') + tags_list = [] + result = "" + for tag in _tags_starting_with(prefix, request.user): + if not tag.name in tags_list: + result += "\n" + tag.name + tags_list.append(tag.name) + dict_result = {"matches": tags_list} + return JSONResponse(dict_result, callback) # ==================== # = Shelf management = @@ -334,24 +447,24 @@ def book_sets(request, slug): book = get_object_or_404(models.Book, slug=slug) user_sets = models.Tag.objects.filter(category='set', user=request.user) book_sets = book.tags.filter(category='set', user=request.user) - + if not request.user.is_authenticated(): return HttpResponse(_('

To maintain your shelves you need to be logged in.

')) - + if request.method == 'POST': form = forms.ObjectSetsForm(book, request.user, request.POST) if form.is_valid(): old_shelves = list(book.tags.filter(category='set')) new_shelves = [models.Tag.objects.get(pk=id) for id in form.cleaned_data['set_ids']] - + for shelf in [shelf for shelf in old_shelves if shelf not in new_shelves]: - shelf.book_count -= 1 + shelf.book_count = None shelf.save() - + for shelf in [shelf for shelf in new_shelves if shelf not in old_shelves]: - shelf.book_count += 1 + shelf.book_count = None shelf.save() - + book.tags = new_shelves + list(book.tags.filter(~Q(category='set') | ~Q(user=request.user))) if request.is_ajax(): return HttpResponse(_('

Shelves were sucessfully saved.

')) @@ -360,7 +473,7 @@ def book_sets(request, slug): else: form = forms.ObjectSetsForm(book, request.user) new_set_form = forms.NewSetForm() - + return render_to_response('catalogue/book_sets.html', locals(), context_instance=RequestContext(request)) @@ -371,11 +484,11 @@ def book_sets(request, slug): def remove_from_shelf(request, shelf, book): book = get_object_or_404(models.Book, slug=book) shelf = get_object_or_404(models.Tag, slug=shelf, category='set', user=request.user) - + if shelf in book.tags: models.Tag.objects.remove_tag(book, shelf) - shelf.book_count -= 1 + shelf.book_count = None shelf.save() return HttpResponse(_('Book was successfully removed from the shelf')) @@ -401,25 +514,30 @@ def download_shelf(request, slug): """" Create a ZIP archive on disk and transmit it in chunks of 8KB, without loading the whole file into memory. A similar approach can - be used for large dynamic PDF files. + be used for large dynamic PDF files. """ shelf = get_object_or_404(models.Tag, slug=slug, category='set') - + formats = [] form = forms.DownloadFormatsForm(request.GET) if form.is_valid(): formats = form.cleaned_data['formats'] if len(formats) == 0: - formats = ['pdf', 'odt', 'txt', 'mp3', 'ogg'] - + formats = ['pdf', 'epub', 'odt', 'txt', 'mp3', 'ogg'] + # Create a ZIP archive temp = tempfile.TemporaryFile() archive = zipfile.ZipFile(temp, 'w') - + + already = set() for book in collect_books(models.Book.tagged.with_all(shelf)): if 'pdf' in formats and book.pdf_file: filename = book.pdf_file.path archive.write(filename, str('%s.pdf' % book.slug)) + if book.root_ancestor not in already and 'epub' in formats and book.root_ancestor.epub_file: + filename = book.root_ancestor.epub_file.path + archive.write(filename, str('%s.epub' % book.root_ancestor.slug)) + already.add(book.root_ancestor) if 'odt' in formats and book.odt_file: filename = book.odt_file.path archive.write(filename, str('%s.odt' % book.slug)) @@ -433,11 +551,11 @@ def download_shelf(request, slug): filename = book.ogg_file.path archive.write(filename, str('%s.ogg' % book.slug)) archive.close() - + response = HttpResponse(content_type='application/zip', mimetype='application/x-zip-compressed') response['Content-Disposition'] = 'attachment; filename=%s.zip' % shelf.sort_key response['Content-Length'] = temp.tell() - + temp.seek(0) response.write(temp.read()) return response @@ -450,11 +568,13 @@ def shelf_book_formats(request, shelf): """ shelf = get_object_or_404(models.Tag, slug=shelf, category='set') - formats = {'pdf': False, 'odt': False, 'txt': False, 'mp3': False, 'ogg': False} - + formats = {'pdf': False, 'epub': False, 'odt': False, 'txt': False, 'mp3': False, 'ogg': False} + for book in collect_books(models.Book.tagged.with_all(shelf)): if book.pdf_file: formats['pdf'] = True + if book.root_ancestor.epub_file: + formats['epub'] = True if book.odt_file: formats['odt'] = True if book.txt_file: @@ -518,7 +638,7 @@ def register(request): if registration_form.is_valid(): user = registration_form.save() user = auth.authenticate( - username=registration_form.cleaned_data['username'], + username=registration_form.cleaned_data['username'], password=registration_form.cleaned_data['password1'] ) auth.login(request, user) @@ -550,7 +670,6 @@ def import_book(request): info = sys.exc_info() exception = pprint.pformat(info[1]) tb = '\n'.join(traceback.format_tb(info[2])) - _('Today is %(month)s, %(day)s.') % {'month': m, 'day': d} return HttpResponse(_("An error occurred: %(exception)s\n\n%(tb)s") % {'exception':exception, 'tb':tb}, mimetype='text/plain') return HttpResponse(_("Book imported successfully")) else: