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