Added some tests for book imports.
[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 tempfile
6 import zipfile
7 import sys
8 import pprint
9 import traceback
10 import re
11
12 from django.conf import settings
13 from django.template import RequestContext
14 from django.shortcuts import render_to_response, get_object_or_404
15 from django.http import HttpResponse, HttpResponseRedirect, Http404
16 from django.core.urlresolvers import reverse
17 from django.db.models import Q
18 from django.contrib.auth.decorators import login_required, user_passes_test
19 from django.utils.datastructures import SortedDict
20 from django.views.decorators.http import require_POST
21 from django.contrib import auth
22 from django.contrib.auth.forms import UserCreationForm, AuthenticationForm
23 from django.utils import simplejson
24 from django.utils.functional import Promise
25 from django.utils.encoding import force_unicode
26 from django.utils.http import urlquote_plus
27 from django.views.decorators import cache
28 from django.utils.translation import ugettext as _
29 from django.views.generic.list_detail import object_list
30
31 from catalogue import models
32 from catalogue import forms
33 from catalogue.utils import split_tags
34 from newtagging import views as newtagging_views
35
36
37 staff_required = user_passes_test(lambda user: user.is_staff)
38
39
40 class LazyEncoder(simplejson.JSONEncoder):
41     def default(self, obj):
42         if isinstance(obj, Promise):
43             return force_unicode(obj)
44         return obj
45
46
47 def main_page(request):
48     if request.user.is_authenticated():
49         shelves = models.Tag.objects.filter(category='set', user=request.user)
50         new_set_form = forms.NewSetForm()
51     extra_where = "NOT catalogue_tag.category = 'set'"
52     tags = models.Tag.objects.usage_for_model(models.Book, counts=True, extra={'where': [extra_where]})
53     fragment_tags = models.Tag.objects.usage_for_model(models.Fragment, counts=True,
54         extra={'where': ["catalogue_tag.category = 'theme'"] + [extra_where]})
55     categories = split_tags(tags)
56
57     form = forms.SearchForm()
58     return render_to_response('catalogue/main_page.html', locals(),
59         context_instance=RequestContext(request))
60
61
62 def book_list(request):
63     books = models.Book.objects.all()
64     form = forms.SearchForm()
65
66     books_by_first_letter = SortedDict()
67     for book in books:
68         books_by_first_letter.setdefault(book.title[0], []).append(book)
69
70     return render_to_response('catalogue/book_list.html', locals(),
71         context_instance=RequestContext(request))
72
73
74 def tagged_object_list(request, tags=''):
75     # Prevent DoS attacks on our database
76     if len(tags.split('/')) > 6:
77         raise Http404
78
79     try:
80         tags = models.Tag.get_tag_list(tags)
81     except models.Tag.DoesNotExist:
82         raise Http404
83
84     if len([tag for tag in tags if tag.category == 'book']):
85         raise Http404
86
87     theme_is_set = [tag for tag in tags if tag.category == 'theme']
88     shelf_is_set = len(tags) == 1 and tags[0].category == 'set'
89     my_shelf_is_set = shelf_is_set and request.user.is_authenticated() and request.user == tags[0].user
90
91     objects = only_author = pd_counter = categories = None
92
93     if theme_is_set:
94         shelf_tags = [tag for tag in tags if tag.category == 'set']
95         fragment_tags = [tag for tag in tags if tag.category != 'set']
96         fragments = models.Fragment.tagged.with_all(fragment_tags)
97
98         if shelf_tags:
99             books = models.Book.tagged.with_all(shelf_tags).order_by()
100             l_tags = [models.Tag.objects.get(slug='l-' + book.slug) for book in books]
101             fragments = models.Fragment.tagged.with_any(l_tags, fragments)
102
103         # newtagging goes crazy if we just try:
104         #related_tags = models.Tag.objects.usage_for_queryset(fragments, counts=True, 
105         #                    extra={'where': ["catalogue_tag.category != 'book'"]})
106         fragment_keys = [fragment.pk for fragment in fragments]
107         if fragment_keys:
108             related_tags = models.Fragment.tags.usage(counts=True,
109                                 filters={'pk__in': fragment_keys},
110                                 extra={'where': ["catalogue_tag.category != 'book'"]})
111             related_tags = (tag for tag in related_tags if tag not in fragment_tags)
112             categories = split_tags(related_tags)
113
114             objects = fragments
115     else:
116         books = models.Book.tagged.with_all(tags).order_by()
117         l_tags = [models.Tag.objects.get(slug='l-' + book.slug) for book in books]
118         book_keys = [book.pk for book in books]
119         # newtagging goes crazy if we just try:
120         #related_tags = models.Tag.objects.usage_for_queryset(books, counts=True, 
121         #                    extra={'where': ["catalogue_tag.category NOT IN ('set', 'book', 'theme')"]})
122         if book_keys:
123             related_tags = models.Book.tags.usage(counts=True,
124                                 filters={'pk__in': book_keys},
125                                 extra={'where': ["catalogue_tag.category NOT IN ('set', 'book', 'theme')"]})
126             categories = split_tags(related_tags)
127
128             fragment_keys = [fragment.pk for fragment in models.Fragment.tagged.with_any(l_tags)]
129             if fragment_keys:
130                 categories['theme'] = models.Fragment.tags.usage(counts=True,
131                                     filters={'pk__in': fragment_keys},
132                                     extra={'where': ["catalogue_tag.category = 'theme'"]})
133
134             books = books.exclude(parent__in=book_keys)
135             objects = books
136
137     if not objects:
138         only_author = len(tags) == 1 and tags[0].category == 'author'
139         pd_counter = only_author and tags[0].goes_to_pd()
140         objects = models.Book.objects.none()
141
142     return object_list(
143         request,
144         objects,
145         template_name='catalogue/tagged_object_list.html',
146         extra_context={
147             'categories': categories,
148             'shelf_is_set': shelf_is_set,
149             'only_author': only_author,
150             'pd_counter': pd_counter,
151             'user_is_owner': my_shelf_is_set,
152             'formats_form': forms.DownloadFormatsForm(),
153
154             'tags': tags,
155         }
156     )
157
158
159 def book_fragments(request, book_slug, theme_slug):
160     book = get_object_or_404(models.Book, slug=book_slug)
161     book_tag = get_object_or_404(models.Tag, slug='l-' + book_slug)
162     theme = get_object_or_404(models.Tag, slug=theme_slug)
163     fragments = models.Fragment.tagged.with_all([book_tag, theme])
164
165     form = forms.SearchForm()
166     return render_to_response('catalogue/book_fragments.html', locals(),
167         context_instance=RequestContext(request))
168
169
170 def book_detail(request, slug):
171     try:
172         book = models.Book.objects.get(slug=slug)
173     except models.Book.DoesNotExist:
174         return book_stub_detail(request, slug)
175
176     book_tag = get_object_or_404(models.Tag, slug='l-' + slug)
177     tags = list(book.tags.filter(~Q(category='set')))
178     categories = split_tags(tags)
179     book_children = book.children.all().order_by('parent_number')
180     extra_where = "catalogue_tag.category = 'theme'"
181     book_themes = models.Tag.objects.related_for_model(book_tag, models.Fragment, counts=True, extra={'where': [extra_where]})
182     extra_info = book.get_extra_info_value()
183
184     form = forms.SearchForm()
185     return render_to_response('catalogue/book_detail.html', locals(),
186         context_instance=RequestContext(request))
187
188
189 def book_stub_detail(request, slug):
190     book = get_object_or_404(models.BookStub, slug=slug)
191     pd_counter = book.pd
192     form = forms.SearchForm()
193
194     return render_to_response('catalogue/book_stub_detail.html', locals(),
195         context_instance=RequestContext(request))
196
197
198 def book_text(request, slug):
199     book = get_object_or_404(models.Book, slug=slug)
200     book_themes = {}
201     for fragment in book.fragments.all():
202         for theme in fragment.tags.filter(category='theme'):
203             book_themes.setdefault(theme, []).append(fragment)
204
205     book_themes = book_themes.items()
206     book_themes.sort(key=lambda s: s[0].sort_key)
207     return render_to_response('catalogue/book_text.html', locals(),
208         context_instance=RequestContext(request))
209
210
211 # ==========
212 # = Search =
213 # ==========
214
215 def _no_diacritics_regexp(query):
216     """ returns a regexp for searching for a query without diacritics
217     
218     should be locale-aware """
219     names = {
220         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',
221         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'
222         }
223     def repl(m):
224         l = m.group()
225         return u"(%s)" % '|'.join(names[l])
226     return re.sub(u'[%s]' % (u''.join(names.keys())), repl, query)
227
228 def unicode_re_escape(query):
229     """ Unicode-friendly version of re.escape """
230     return re.sub('(?u)(\W)', r'\\\1', query)
231
232 def _word_starts_with(name, prefix):
233     """returns a Q object getting models having `name` contain a word
234     starting with `prefix`
235     
236     We define word characters as alphanumeric and underscore, like in JS.
237     
238     Works for MySQL, PostgreSQL, Oracle.
239     For SQLite, _sqlite* version is substituted for this.
240     """
241     kwargs = {}
242
243     prefix = _no_diacritics_regexp(unicode_re_escape(prefix))
244     # can't use [[:<:]] (word start), 
245     # but we want both `xy` and `(xy` to catch `(xyz)`
246     kwargs['%s__iregex' % name] = u"(^|[^[:alnum:]_])%s" % prefix
247
248     return Q(**kwargs)
249
250
251 def _sqlite_word_starts_with(name, prefix):
252     """ version of _word_starts_with for SQLite 
253     
254     SQLite in Django uses Python re module
255     """
256     kwargs = {}
257     prefix = _no_diacritics_regexp(unicode_re_escape(prefix))
258     kwargs['%s__iregex' % name] = ur"(^|(?<=[^\wąćęłńóśźżĄĆĘŁŃÓŚŹŻ]))%s" % prefix
259     return Q(**kwargs)
260
261
262 if settings.DATABASE_ENGINE == 'sqlite3':
263     _word_starts_with = _sqlite_word_starts_with
264
265
266 def _tags_starting_with(prefix, user=None):
267     prefix = prefix.lower()
268     book_stubs = models.BookStub.objects.filter(_word_starts_with('title', prefix))
269     books = models.Book.objects.filter(_word_starts_with('title', prefix))
270     book_stubs = filter(lambda x: x not in books, book_stubs)
271     tags = models.Tag.objects.filter(_word_starts_with('name', prefix))
272     if user and user.is_authenticated():
273         tags = tags.filter(~Q(category='book') & (~Q(category='set') | Q(user=user)))
274     else:
275         tags = tags.filter(~Q(category='book') & ~Q(category='set'))
276
277     return list(books) + list(tags) + list(book_stubs)
278
279
280 def _get_result_link(match, tag_list):
281     if isinstance(match, models.Book) or isinstance(match, models.BookStub):
282         return match.get_absolute_url()
283     else:
284         return reverse('catalogue.views.tagged_object_list',
285             kwargs={'tags': '/'.join(tag.slug for tag in tag_list + [match])}
286         )
287
288 def _get_result_type(match):
289     if isinstance(match, models.Book) or isinstance(match, models.BookStub):
290         type = 'book'
291     else:
292         type = match.category
293     return dict(models.TAG_CATEGORIES)[type]
294
295
296
297 def find_best_matches(query, user=None):
298     """ Finds a Book, Tag or Bookstub best matching a query.
299     
300     Returns a with:
301       - zero elements when nothing is found,
302       - one element when a best result is found,
303       - more then one element on multiple exact matches
304     
305     Raises a ValueError on too short a query.
306     """
307
308     query = query.lower()
309     if len(query) < 2:
310         raise ValueError("query must have at least two characters")
311
312     result = tuple(_tags_starting_with(query, user))
313     exact_matches = tuple(res for res in result if res.name.lower() == query)
314     if exact_matches:
315         return exact_matches
316     else:
317         return result[:1]
318
319
320 def search(request):
321     tags = request.GET.get('tags', '')
322     prefix = request.GET.get('q', '')
323
324     try:
325         tag_list = models.Tag.get_tag_list(tags)
326     except:
327         tag_list = []
328
329     try:
330         result = find_best_matches(prefix, request.user)
331     except ValueError:
332         return render_to_response('catalogue/search_too_short.html', {'tags':tag_list, 'prefix':prefix},
333             context_instance=RequestContext(request))
334
335     if len(result) == 1:
336         return HttpResponseRedirect(_get_result_link(result[0], tag_list))
337     elif len(result) > 1:
338         return render_to_response('catalogue/search_multiple_hits.html',
339             {'tags':tag_list, 'prefix':prefix, 'results':((x, _get_result_link(x, tag_list), _get_result_type(x)) for x in result)},
340             context_instance=RequestContext(request))
341     else:
342         return render_to_response('catalogue/search_no_hits.html', {'tags':tag_list, 'prefix':prefix},
343             context_instance=RequestContext(request))
344
345
346 def tags_starting_with(request):
347     prefix = request.GET.get('q', '')
348     # Prefix must have at least 2 characters
349     if len(prefix) < 2:
350         return HttpResponse('')
351
352     return HttpResponse('\n'.join(tag.name for tag in _tags_starting_with(prefix, request.user)))
353
354
355 # ====================
356 # = Shelf management =
357 # ====================
358 @login_required
359 @cache.never_cache
360 def user_shelves(request):
361     shelves = models.Tag.objects.filter(category='set', user=request.user)
362     new_set_form = forms.NewSetForm()
363     return render_to_response('catalogue/user_shelves.html', locals(),
364             context_instance=RequestContext(request))
365
366 @cache.never_cache
367 def book_sets(request, slug):
368     book = get_object_or_404(models.Book, slug=slug)
369     user_sets = models.Tag.objects.filter(category='set', user=request.user)
370     book_sets = book.tags.filter(category='set', user=request.user)
371
372     if not request.user.is_authenticated():
373         return HttpResponse(_('<p>To maintain your shelves you need to be logged in.</p>'))
374
375     if request.method == 'POST':
376         form = forms.ObjectSetsForm(book, request.user, request.POST)
377         if form.is_valid():
378             old_shelves = list(book.tags.filter(category='set'))
379             new_shelves = [models.Tag.objects.get(pk=id) for id in form.cleaned_data['set_ids']]
380
381             for shelf in [shelf for shelf in old_shelves if shelf not in new_shelves]:
382                 shelf.book_count -= 1
383                 shelf.save()
384
385             for shelf in [shelf for shelf in new_shelves if shelf not in old_shelves]:
386                 shelf.book_count += 1
387                 shelf.save()
388
389             book.tags = new_shelves + list(book.tags.filter(~Q(category='set') | ~Q(user=request.user)))
390             if request.is_ajax():
391                 return HttpResponse(_('<p>Shelves were sucessfully saved.</p>'))
392             else:
393                 return HttpResponseRedirect('/')
394     else:
395         form = forms.ObjectSetsForm(book, request.user)
396         new_set_form = forms.NewSetForm()
397
398     return render_to_response('catalogue/book_sets.html', locals(),
399         context_instance=RequestContext(request))
400
401
402 @login_required
403 @require_POST
404 @cache.never_cache
405 def remove_from_shelf(request, shelf, book):
406     book = get_object_or_404(models.Book, slug=book)
407     shelf = get_object_or_404(models.Tag, slug=shelf, category='set', user=request.user)
408
409     if shelf in book.tags:
410         models.Tag.objects.remove_tag(book, shelf)
411
412         shelf.book_count -= 1
413         shelf.save()
414
415         return HttpResponse(_('Book was successfully removed from the shelf'))
416     else:
417         return HttpResponse(_('This book is not on the shelf'))
418
419
420 def collect_books(books):
421     """
422     Returns all real books in collection.
423     """
424     result = []
425     for book in books:
426         if len(book.children.all()) == 0:
427             result.append(book)
428         else:
429             result += collect_books(book.children.all())
430     return result
431
432
433 @cache.never_cache
434 def download_shelf(request, slug):
435     """"
436     Create a ZIP archive on disk and transmit it in chunks of 8KB,
437     without loading the whole file into memory. A similar approach can
438     be used for large dynamic PDF files.                                        
439     """
440     shelf = get_object_or_404(models.Tag, slug=slug, category='set')
441
442     formats = []
443     form = forms.DownloadFormatsForm(request.GET)
444     if form.is_valid():
445         formats = form.cleaned_data['formats']
446     if len(formats) == 0:
447         formats = ['pdf', 'odt', 'txt', 'mp3', 'ogg']
448
449     # Create a ZIP archive
450     temp = tempfile.TemporaryFile()
451     archive = zipfile.ZipFile(temp, 'w')
452
453     for book in collect_books(models.Book.tagged.with_all(shelf)):
454         if 'pdf' in formats and book.pdf_file:
455             filename = book.pdf_file.path
456             archive.write(filename, str('%s.pdf' % book.slug))
457         if 'odt' in formats and book.odt_file:
458             filename = book.odt_file.path
459             archive.write(filename, str('%s.odt' % book.slug))
460         if 'txt' in formats and book.txt_file:
461             filename = book.txt_file.path
462             archive.write(filename, str('%s.txt' % book.slug))
463         if 'mp3' in formats and book.mp3_file:
464             filename = book.mp3_file.path
465             archive.write(filename, str('%s.mp3' % book.slug))
466         if 'ogg' in formats and book.ogg_file:
467             filename = book.ogg_file.path
468             archive.write(filename, str('%s.ogg' % book.slug))
469     archive.close()
470
471     response = HttpResponse(content_type='application/zip', mimetype='application/x-zip-compressed')
472     response['Content-Disposition'] = 'attachment; filename=%s.zip' % shelf.sort_key
473     response['Content-Length'] = temp.tell()
474
475     temp.seek(0)
476     response.write(temp.read())
477     return response
478
479
480 @cache.never_cache
481 def shelf_book_formats(request, shelf):
482     """"
483     Returns a list of formats of books in shelf.
484     """
485     shelf = get_object_or_404(models.Tag, slug=shelf, category='set')
486
487     formats = {'pdf': False, 'odt': False, 'txt': False, 'mp3': False, 'ogg': False}
488
489     for book in collect_books(models.Book.tagged.with_all(shelf)):
490         if book.pdf_file:
491             formats['pdf'] = True
492         if book.odt_file:
493             formats['odt'] = True
494         if book.txt_file:
495             formats['txt'] = True
496         if book.mp3_file:
497             formats['mp3'] = True
498         if book.ogg_file:
499             formats['ogg'] = True
500
501     return HttpResponse(LazyEncoder().encode(formats))
502
503
504 @login_required
505 @require_POST
506 @cache.never_cache
507 def new_set(request):
508     new_set_form = forms.NewSetForm(request.POST)
509     if new_set_form.is_valid():
510         new_set = new_set_form.save(request.user)
511
512         if request.is_ajax():
513             return HttpResponse(_('<p>Shelf <strong>%s</strong> was successfully created</p>') % new_set)
514         else:
515             return HttpResponseRedirect('/')
516
517     return HttpResponseRedirect('/')
518
519
520 @login_required
521 @require_POST
522 @cache.never_cache
523 def delete_shelf(request, slug):
524     user_set = get_object_or_404(models.Tag, slug=slug, category='set', user=request.user)
525     user_set.delete()
526
527     if request.is_ajax():
528         return HttpResponse(_('<p>Shelf <strong>%s</strong> was successfully removed</p>') % user_set.name)
529     else:
530         return HttpResponseRedirect('/')
531
532
533 # ==================
534 # = Authentication =
535 # ==================
536 @require_POST
537 @cache.never_cache
538 def login(request):
539     form = AuthenticationForm(data=request.POST, prefix='login')
540     if form.is_valid():
541         auth.login(request, form.get_user())
542         response_data = {'success': True, 'errors': {}}
543     else:
544         response_data = {'success': False, 'errors': form.errors}
545     return HttpResponse(LazyEncoder(ensure_ascii=False).encode(response_data))
546
547
548 @require_POST
549 @cache.never_cache
550 def register(request):
551     registration_form = UserCreationForm(request.POST, prefix='registration')
552     if registration_form.is_valid():
553         user = registration_form.save()
554         user = auth.authenticate(
555             username=registration_form.cleaned_data['username'],
556             password=registration_form.cleaned_data['password1']
557         )
558         auth.login(request, user)
559         response_data = {'success': True, 'errors': {}}
560     else:
561         response_data = {'success': False, 'errors': registration_form.errors}
562     return HttpResponse(LazyEncoder(ensure_ascii=False).encode(response_data))
563
564
565 @cache.never_cache
566 def logout_then_redirect(request):
567     auth.logout(request)
568     return HttpResponseRedirect(urlquote_plus(request.GET.get('next', '/'), safe='/?='))
569
570
571
572 # =========
573 # = Admin =
574 # =========
575 @login_required
576 @staff_required
577 def import_book(request):
578     """docstring for import_book"""
579     book_import_form = forms.BookImportForm(request.POST, request.FILES)
580     if book_import_form.is_valid():
581         try:
582             book_import_form.save()
583         except:
584             info = sys.exc_info()
585             exception = pprint.pformat(info[1])
586             tb = '\n'.join(traceback.format_tb(info[2]))
587             _('Today is %(month)s, %(day)s.') % {'month': m, 'day': d}
588             return HttpResponse(_("An error occurred: %(exception)s\n\n%(tb)s") % {'exception':exception, 'tb':tb}, mimetype='text/plain')
589         return HttpResponse(_("Book imported successfully"))
590     else:
591         return HttpResponse(_("Error importing file: %r") % book_import_form.errors)
592
593
594
595 def clock(request):
596     """ Provides server time for jquery.countdown,
597     in a format suitable for Date.parse()
598     """
599     from datetime import datetime
600     return HttpResponse(datetime.now().strftime('%Y/%m/%d %H:%M:%S'))