fix bug with related tags and object display when tag spanned both books and pictures
[wolnelektury.git] / apps / catalogue / views.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 import re
6 import itertools
7
8 from django.conf import settings
9 from django.core.cache import get_cache
10 from django.template import RequestContext
11 from django.template.loader import render_to_string
12 from django.shortcuts import render_to_response, get_object_or_404, redirect
13 from django.http import HttpResponse, HttpResponseRedirect, Http404, HttpResponsePermanentRedirect
14 from django.core.urlresolvers import reverse
15 from django.db.models import Q
16 from django.contrib.auth.decorators import login_required, user_passes_test
17 from django.utils.datastructures import SortedDict
18 from django.utils.http import urlquote_plus
19 from django.utils import translation
20 from django.utils.translation import get_language, ugettext as _, ugettext_lazy
21 from django.views.decorators.vary import vary_on_headers
22
23 from ajaxable.utils import JSONResponse, AjaxableFormView
24 from catalogue import models
25 from catalogue import forms
26 from catalogue.utils import split_tags, MultiQuerySet, SortedMultiQuerySet
27 from catalogue.templatetags.catalogue_tags import tag_list, collection_list
28 from pdcounter import models as pdcounter_models
29 from pdcounter import views as pdcounter_views
30 from suggest.forms import PublishingSuggestForm
31 from picture.models import Picture, PictureArea
32 from picture.views import picture_list_thumb
33 import logging
34 staff_required = user_passes_test(lambda user: user.is_staff)
35 permanent_cache = get_cache('permanent')
36
37
38 @vary_on_headers('X-Requested-With')
39 def catalogue(request):
40     cache_key='catalogue.catalogue/' + get_language()
41     output = permanent_cache.get(cache_key)
42
43     if output is None:
44         tags = models.Tag.objects.exclude(
45             category__in=('set', 'book')).exclude(book_count=0, picture_count=0)
46         tags = list(tags)
47         for tag in tags:
48             tag.count = tag.book_count + tag.picture_count
49         categories = split_tags(tags)
50         fragment_tags = categories.get('theme', [])
51         collections = models.Collection.objects.all()
52
53         render_tag_list = lambda x: render_to_string(
54             'catalogue/tag_list.html', tag_list(x))
55         has_pictures = lambda x: filter(lambda y: y.picture_count>0, x)
56         has_books = lambda x: filter(lambda y: y.book_count>0, x)
57         def render_split(tags):
58             with_books = has_books(tags)
59             with_pictures = has_pictures(tags)
60             ctx = {}
61             if with_books:
62                 ctx['books'] = render_tag_list(with_books)
63             if with_pictures:
64                 ctx['pictures'] = render_tag_list(with_pictures)
65             return render_to_string('catalogue/tag_list_split.html', ctx)
66
67         output = {'theme': {}}
68         output['theme'] = render_tag_list(fragment_tags)
69         for category, tags in categories.items():
70             if category in ('author', 'theme'):
71                 output[category] = render_tag_list(tags)
72             else:
73                 output[category] = render_split(tags)
74             
75             
76         output['collections'] = render_to_string(
77             'catalogue/collection_list.html', collection_list(collections))
78         permanent_cache.set(cache_key, output)
79     if request.is_ajax():
80         return JSONResponse(output)
81     else:
82         return render_to_response('catalogue/catalogue.html', locals(),
83             context_instance=RequestContext(request))
84
85
86 def book_list(request, filter=None, get_filter=None,
87         template_name='catalogue/book_list.html',
88         nav_template_name='catalogue/snippets/book_list_nav.html',
89         list_template_name='catalogue/snippets/book_list.html',
90         cache_key='catalogue.book_list',
91         context=None,
92         ):
93     """ generates a listing of all books, optionally filtered with a test function """
94     cache_key = "%s/%s" % (cache_key, get_language())
95     cached = permanent_cache.get(cache_key)
96     if cached is not None:
97         rendered_nav, rendered_book_list = cached
98     else:
99         if get_filter:
100             filter = get_filter()
101         books_by_author, orphans, books_by_parent = models.Book.book_list(filter)
102         books_nav = SortedDict()
103         for tag in books_by_author:
104             if books_by_author[tag]:
105                 books_nav.setdefault(tag.sort_key[0], []).append(tag)
106         rendered_nav = render_to_string(nav_template_name, locals())
107         rendered_book_list = render_to_string(list_template_name, locals())
108         permanent_cache.set(cache_key, (rendered_nav, rendered_book_list))
109     return render_to_response(template_name, locals(),
110         context_instance=RequestContext(request))
111
112
113 def audiobook_list(request):
114     return book_list(request, Q(media__type='mp3') | Q(media__type='ogg'),
115                      template_name='catalogue/audiobook_list.html',
116                      list_template_name='catalogue/snippets/audiobook_list.html',
117                      cache_key='catalogue.audiobook_list')
118
119
120 def daisy_list(request):
121     return book_list(request, Q(media__type='daisy'),
122                      template_name='catalogue/daisy_list.html',
123                      cache_key='catalogue.daisy_list')
124
125
126 def collection(request, slug):
127     coll = get_object_or_404(models.Collection, slug=slug)
128     if coll.kind == 'book':
129         view = book_list
130         tmpl = "catalogue/collection.html"
131     elif coll.kind == 'picture':
132         view = picture_list_thumb
133         tmpl = "picture/collection.html"
134     else:
135         raise ValueError('How do I show this kind of collection? %s' % coll.kind)
136     return view(request, get_filter=coll.get_query,
137                      template_name=tmpl,
138                      cache_key='catalogue.collection:%s' % coll.slug,
139                      context={'collection': coll})
140
141
142 def differentiate_tags(request, tags, ambiguous_slugs):
143     beginning = '/'.join(tag.url_chunk for tag in tags)
144     unparsed = '/'.join(ambiguous_slugs[1:])
145     options = []
146     for tag in models.Tag.objects.exclude(category='book').filter(slug=ambiguous_slugs[0]):
147         options.append({
148             'url_args': '/'.join((beginning, tag.url_chunk, unparsed)).strip('/'),
149             'tags': [tag]
150         })
151     return render_to_response('catalogue/differentiate_tags.html',
152                 {'tags': tags, 'options': options, 'unparsed': ambiguous_slugs[1:]},
153                 context_instance=RequestContext(request))
154
155
156 # TODO: Rewrite this hellish piece of code which tries to do everything
157 def tagged_object_list(request, tags=''):
158     # preliminary tests and conditions
159     try:
160         tags = models.Tag.get_tag_list(tags)
161     except models.Tag.DoesNotExist:
162         # Perhaps the user is asking about an author in Public Domain
163         # counter (they are not represented in tags)
164         chunks = tags.split('/')
165         if len(chunks) == 2 and chunks[0] == 'autor':
166             return pdcounter_views.author_detail(request, chunks[1])
167         else:
168             raise Http404
169     except models.Tag.MultipleObjectsReturned, e:
170         # Ask the user to disambiguate
171         return differentiate_tags(request, e.tags, e.ambiguous_slugs)
172     except models.Tag.UrlDeprecationWarning, e:
173         return HttpResponsePermanentRedirect(reverse('tagged_object_list', args=['/'.join(tag.url_chunk for tag in e.tags)]))
174
175     try:
176         if len(tags) > settings.MAX_TAG_LIST:
177             raise Http404
178     except AttributeError:
179         pass
180
181     if len([tag for tag in tags if tag.category == 'book']):
182         raise Http404
183
184     # beginning of digestion
185     theme_is_set = [tag for tag in tags if tag.category == 'theme']
186     shelf_is_set = [tag for tag in tags if tag.category == 'set']
187     only_shelf = shelf_is_set and len(tags) == 1
188     only_my_shelf = only_shelf and request.user.is_authenticated() and request.user == tags[0].user
189
190
191     objects = only_author = None
192     categories = {}
193     object_queries = []
194
195     if theme_is_set:
196         shelf_tags = [tag for tag in tags if tag.category == 'set']
197         fragment_tags = [tag for tag in tags if tag.category != 'set']
198         fragments = models.Fragment.tagged.with_all(fragment_tags)
199         areas = PictureArea.tagged.with_all(fragment_tags)
200
201         if shelf_tags:
202             books = models.Book.tagged.with_all(shelf_tags).order_by()
203             l_tags = models.Tag.objects.filter(category='book',
204                 slug__in=[book.book_tag_slug() for book in books.iterator()])
205             fragments = models.Fragment.tagged.with_any(l_tags, fragments)
206
207         # newtagging goes crazy if we just try:
208         #related_tags = models.Tag.objects.usage_for_queryset(fragments, counts=True,
209         #                    extra={'where': ["catalogue_tag.category != 'book'"]})
210
211         related_tags = []
212
213         fragment_keys = [fragment.pk for fragment in fragments.iterator()]
214         if fragment_keys:
215             related_tags = models.Fragment.tags.usage(counts=True,
216                                 filters={'pk__in': fragment_keys},
217                                 extra={'where': ["catalogue_tag.category != 'book'"]})
218             related_tags = (tag for tag in related_tags if tag not in fragment_tags)
219             categories = split_tags(related_tags, categories)
220
221         object_queries.insert(0, fragments)
222
223         area_keys = [area.pk for area in areas.iterator()]
224         if area_keys:
225             related_tags = PictureArea.tags.usage(counts=True,
226                                                          filters={'pk__in': area_keys})
227             related_tags = (tag for tag in related_tags if tag not in fragment_tags)     
228                                                       
229             categories = split_tags(related_tags, categories)
230
231         # we want the Pictures to go first
232         object_queries.insert(0, areas)
233         objects = MultiQuerySet(*object_queries)
234     else:
235         if shelf_is_set:
236             books = models.Book.tagged.with_all(tags).order_by('sort_key_author')
237         else:
238             books = models.Book.tagged_top_level(tags).order_by('sort_key_author')
239
240         pictures = Picture.tagged.with_all(tags).order_by('sort_key_author')
241             
242         related_counts = {}
243         if books.count() > 0:
244             # get related tags from `tag_counter` and `theme_counter`
245             tags_pks = [tag.pk for tag in tags]
246             for book in books:
247                 for tag_pk, value in itertools.chain(book.tag_counter.iteritems(), book.theme_counter.iteritems()):
248                     if tag_pk in tags_pks:
249                         continue
250                     related_counts[tag_pk] = related_counts.get(tag_pk, 0) + value
251
252         if pictures.count() > 0:
253             tags_pks = [tag.pk for tag in tags]
254             for picture in pictures:
255                 for tag_pk, value in itertools.chain(picture.tag_counter.iteritems(), picture.theme_counter.iteritems()):
256                     if tag_pk in tags_pks:
257                         continue
258                     related_counts[tag_pk] = related_counts.get(tag_pk, 0) + value
259
260         related_tags = models.Tag.objects.filter(pk__in=related_counts.keys())
261         related_tags = [tag for tag in related_tags if tag not in tags]
262
263         for tag in related_tags:
264             tag.count = related_counts[tag.pk]
265
266         categories = split_tags(related_tags)
267         del related_tags
268
269
270         objects = SortedMultiQuerySet(pictures, books, order_by='sort_key_author')
271
272
273     if not objects:
274         only_author = len(tags) == 1 and tags[0].category == 'author'
275         objects = models.Book.objects.none()
276
277     return render_to_response('catalogue/tagged_object_list.html',
278         {
279             'object_list': objects,
280             'categories': categories,
281             'only_shelf': only_shelf,
282             'only_author': only_author,
283             'only_my_shelf': only_my_shelf,
284             'formats_form': forms.DownloadFormatsForm(),
285             'tags': tags,
286             'theme_is_set': theme_is_set,
287         },
288         context_instance=RequestContext(request))
289
290
291 def book_fragments(request, slug, theme_slug):
292     book = get_object_or_404(models.Book, slug=slug)
293
294     book_tag = book.book_tag()
295     theme = get_object_or_404(models.Tag, slug=theme_slug, category='theme')
296     fragments = models.Fragment.tagged.with_all([book_tag, theme])
297
298     return render_to_response('catalogue/book_fragments.html', locals(),
299         context_instance=RequestContext(request))
300
301
302 def book_detail(request, slug):
303     try:
304         book = models.Book.objects.get(slug=slug)
305     except models.Book.DoesNotExist:
306         return pdcounter_views.book_stub_detail(request, slug)
307
308     book_children = book.children.all().order_by('parent_number', 'sort_key')
309     return render_to_response('catalogue/book_detail.html', locals(),
310         context_instance=RequestContext(request))
311
312
313 def player(request, slug):
314     book = get_object_or_404(models.Book, slug=slug)
315     if not book.has_media('mp3'):
316         raise Http404
317
318     ogg_files = {}
319     for m in book.media.filter(type='ogg').order_by().iterator():
320         ogg_files[m.name] = m
321
322     audiobooks = []
323     have_oggs = True
324     projects = set()
325     for mp3 in book.media.filter(type='mp3').iterator():
326         # ogg files are always from the same project
327         meta = mp3.extra_info
328         project = meta.get('project')
329         if not project:
330             # temporary fallback
331             project = u'CzytamySłuchając'
332
333         projects.add((project, meta.get('funded_by', '')))
334
335         media = {'mp3': mp3}
336
337         ogg = ogg_files.get(mp3.name)
338         if ogg:
339             media['ogg'] = ogg
340         else:
341             have_oggs = False
342         audiobooks.append(media)
343
344     projects = sorted(projects)
345
346     extra_info = book.extra_info
347
348     return render_to_response('catalogue/player.html', locals(),
349         context_instance=RequestContext(request))
350
351
352 def book_text(request, slug):
353     book = get_object_or_404(models.Book, slug=slug)
354
355     if not book.has_html_file():
356         raise Http404
357     related = book.related_info()
358     return render_to_response('catalogue/book_text.html', locals(),
359         context_instance=RequestContext(request))
360
361
362 # ==========
363 # = Search =
364 # ==========
365
366 def _no_diacritics_regexp(query):
367     """ returns a regexp for searching for a query without diacritics
368
369     should be locale-aware """
370     names = {
371         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',
372         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'
373         }
374     def repl(m):
375         l = m.group()
376         return u"(%s)" % '|'.join(names[l])
377     return re.sub(u'[%s]' % (u''.join(names.keys())), repl, query)
378
379 def unicode_re_escape(query):
380     """ Unicode-friendly version of re.escape """
381     return re.sub('(?u)(\W)', r'\\\1', query)
382
383 def _word_starts_with(name, prefix):
384     """returns a Q object getting models having `name` contain a word
385     starting with `prefix`
386
387     We define word characters as alphanumeric and underscore, like in JS.
388
389     Works for MySQL, PostgreSQL, Oracle.
390     For SQLite, _sqlite* version is substituted for this.
391     """
392     kwargs = {}
393
394     prefix = _no_diacritics_regexp(unicode_re_escape(prefix))
395     # can't use [[:<:]] (word start),
396     # but we want both `xy` and `(xy` to catch `(xyz)`
397     kwargs['%s__iregex' % name] = u"(^|[^[:alnum:]_])%s" % prefix
398
399     return Q(**kwargs)
400
401
402 def _word_starts_with_regexp(prefix):
403     prefix = _no_diacritics_regexp(unicode_re_escape(prefix))
404     return ur"(^|(?<=[^\wąćęłńóśźżĄĆĘŁŃÓŚŹŻ]))%s" % prefix
405
406
407 def _sqlite_word_starts_with(name, prefix):
408     """ version of _word_starts_with for SQLite
409
410     SQLite in Django uses Python re module
411     """
412     kwargs = {}
413     kwargs['%s__iregex' % name] = _word_starts_with_regexp(prefix)
414     return Q(**kwargs)
415
416
417 if hasattr(settings, 'DATABASES'):
418     if settings.DATABASES['default']['ENGINE'] == 'django.db.backends.sqlite3':
419         _word_starts_with = _sqlite_word_starts_with
420 elif settings.DATABASE_ENGINE == 'sqlite3':
421     _word_starts_with = _sqlite_word_starts_with
422
423
424 class App():
425     def __init__(self, name, view):
426         self.name = name
427         self._view = view
428         self.lower = name.lower()
429         self.category = 'application'
430     def view(self):
431         return reverse(*self._view)
432
433 _apps = (
434     App(u'Leśmianator', (u'lesmianator', )),
435     )
436
437
438 def _tags_starting_with(prefix, user=None):
439     prefix = prefix.lower()
440     # PD counter
441     book_stubs = pdcounter_models.BookStub.objects.filter(_word_starts_with('title', prefix))
442     authors = pdcounter_models.Author.objects.filter(_word_starts_with('name', prefix))
443
444     books = models.Book.objects.filter(_word_starts_with('title', prefix))
445     tags = models.Tag.objects.filter(_word_starts_with('name', prefix))
446     if user and user.is_authenticated():
447         tags = tags.filter(~Q(category='book') & (~Q(category='set') | Q(user=user)))
448     else:
449         tags = tags.filter(~Q(category='book') & ~Q(category='set'))
450
451     prefix_regexp = re.compile(_word_starts_with_regexp(prefix))
452     return list(books) + list(tags) + [app for app in _apps if prefix_regexp.search(app.lower)] + list(book_stubs) + list(authors)
453
454
455 def _get_result_link(match, tag_list):
456     if isinstance(match, models.Tag):
457         return reverse('catalogue.views.tagged_object_list',
458             kwargs={'tags': '/'.join(tag.url_chunk for tag in tag_list + [match])}
459         )
460     elif isinstance(match, App):
461         return match.view()
462     else:
463         return match.get_absolute_url()
464
465
466 def _get_result_type(match):
467     if isinstance(match, models.Book) or isinstance(match, pdcounter_models.BookStub):
468         type = 'book'
469     else:
470         type = match.category
471     return type
472
473
474 def books_starting_with(prefix):
475     prefix = prefix.lower()
476     return models.Book.objects.filter(_word_starts_with('title', prefix))
477
478
479 def find_best_matches(query, user=None):
480     """ Finds a models.Book, Tag, models.BookStub or Author best matching a query.
481
482     Returns a with:
483       - zero elements when nothing is found,
484       - one element when a best result is found,
485       - more then one element on multiple exact matches
486
487     Raises a ValueError on too short a query.
488     """
489
490     query = query.lower()
491     if len(query) < 2:
492         raise ValueError("query must have at least two characters")
493
494     result = tuple(_tags_starting_with(query, user))
495     # remove pdcounter stuff
496     book_titles = set(match.pretty_title().lower() for match in result
497                       if isinstance(match, models.Book))
498     authors = set(match.name.lower() for match in result
499                   if isinstance(match, models.Tag) and match.category=='author')
500     result = tuple(res for res in result if not (
501                  (isinstance(res, pdcounter_models.BookStub) and res.pretty_title().lower() in book_titles)
502                  or (isinstance(res, pdcounter_models.Author) and res.name.lower() in authors)
503              ))
504
505     exact_matches = tuple(res for res in result if res.name.lower() == query)
506     if exact_matches:
507         return exact_matches
508     else:
509         return tuple(result)[:1]
510
511
512 def search(request):
513     tags = request.GET.get('tags', '')
514     prefix = request.GET.get('q', '')
515
516     try:
517         tag_list = models.Tag.get_tag_list(tags)
518     except:
519         tag_list = []
520
521     try:
522         result = find_best_matches(prefix, request.user)
523     except ValueError:
524         return render_to_response('catalogue/search_too_short.html', {'tags':tag_list, 'prefix':prefix},
525             context_instance=RequestContext(request))
526
527     if len(result) == 1:
528         return HttpResponseRedirect(_get_result_link(result[0], tag_list))
529     elif len(result) > 1:
530         return render_to_response('catalogue/search_multiple_hits.html',
531             {'tags':tag_list, 'prefix':prefix, 'results':((x, _get_result_link(x, tag_list), _get_result_type(x)) for x in result)},
532             context_instance=RequestContext(request))
533     else:
534         form = PublishingSuggestForm(initial={"books": prefix + ", "})
535         return render_to_response('catalogue/search_no_hits.html',
536             {'tags':tag_list, 'prefix':prefix, "pubsuggest_form": form},
537             context_instance=RequestContext(request))
538
539
540 def tags_starting_with(request):
541     prefix = request.GET.get('q', '')
542     # Prefix must have at least 2 characters
543     if len(prefix) < 2:
544         return HttpResponse('')
545     tags_list = []
546     result = ""
547     for tag in _tags_starting_with(prefix, request.user):
548         if not tag.name in tags_list:
549             result += "\n" + tag.name
550             tags_list.append(tag.name)
551     return HttpResponse(result)
552
553 def json_tags_starting_with(request, callback=None):
554     # Callback for JSONP
555     prefix = request.GET.get('q', '')
556     callback = request.GET.get('callback', '')
557     # Prefix must have at least 2 characters
558     if len(prefix) < 2:
559         return HttpResponse('')
560     tags_list = []
561     for tag in _tags_starting_with(prefix, request.user):
562         if not tag.name in tags_list:
563             tags_list.append(tag.name)
564     if request.GET.get('mozhint', ''):
565         result = [prefix, tags_list]
566     else:
567         result = {"matches": tags_list}
568     return JSONResponse(result, callback)
569
570
571 # =========
572 # = Admin =
573 # =========
574 @login_required
575 @staff_required
576 def import_book(request):
577     """docstring for import_book"""
578     book_import_form = forms.BookImportForm(request.POST, request.FILES)
579     if book_import_form.is_valid():
580         try:
581             book_import_form.save()
582         except:
583             import sys
584             import pprint
585             import traceback
586             info = sys.exc_info()
587             exception = pprint.pformat(info[1])
588             tb = '\n'.join(traceback.format_tb(info[2]))
589             return HttpResponse(_("An error occurred: %(exception)s\n\n%(tb)s") % {'exception':exception, 'tb':tb}, mimetype='text/plain')
590         return HttpResponse(_("Book imported successfully"))
591     else:
592         return HttpResponse(_("Error importing file: %r") % book_import_form.errors)
593
594
595 # info views for API
596
597 def book_info(request, id, lang='pl'):
598     book = get_object_or_404(models.Book, id=id)
599     # set language by hand
600     translation.activate(lang)
601     return render_to_response('catalogue/book_info.html', locals(),
602         context_instance=RequestContext(request))
603
604
605 def tag_info(request, id):
606     tag = get_object_or_404(models.Tag, id=id)
607     return HttpResponse(tag.description)
608
609
610 def download_zip(request, format, slug=None):
611     url = None
612     if format in models.Book.ebook_formats:
613         url = models.Book.zip_format(format)
614     elif format in ('mp3', 'ogg') and slug is not None:
615         book = get_object_or_404(models.Book, slug=slug)
616         url = book.zip_audiobooks(format)
617     else:
618         raise Http404('No format specified for zip package')
619     return HttpResponseRedirect(urlquote_plus(settings.MEDIA_URL + url, safe='/?='))
620
621
622 class CustomPDFFormView(AjaxableFormView):
623     form_class = forms.CustomPDFForm
624     title = ugettext_lazy('Download custom PDF')
625     submit = ugettext_lazy('Download')
626     honeypot = True
627
628     def __call__(self, *args, **kwargs):
629         if settings.NO_CUSTOM_PDF:
630             raise Http404('Custom PDF is disabled')
631         return super(CustomPDFFormView, self).__call__(*args, **kwargs)
632
633     def form_args(self, request, obj):
634         """Override to parse view args and give additional args to the form."""
635         return (obj,), {}
636
637     def get_object(self, request, slug, *args, **kwargs):
638         return get_object_or_404(models.Book, slug=slug)
639
640     def context_description(self, request, obj):
641         return obj.pretty_title()