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 import itertools
  12 from operator import itemgetter
  13
  14 from django.conf import settings
  15 from django.template import RequestContext
  16 from django.shortcuts import render_to_response, get_object_or_404
  17 from django.http import HttpResponse, HttpResponseRedirect, Http404
  18 from django.core.urlresolvers import reverse
  19 from django.db.models import Q
  20 from django.contrib.auth.decorators import login_required, user_passes_test
  21 from django.utils.datastructures import SortedDict
  22 from django.views.decorators.http import require_POST
  23 from django.contrib import auth
  24 from django.contrib.auth.forms import UserCreationForm, AuthenticationForm
  25 from django.utils import simplejson
  26 from django.utils.functional import Promise
  27 from django.utils.encoding import force_unicode
  28 from django.utils.http import urlquote_plus
  29 from django.views.decorators import cache
  30 from django.utils.translation import ugettext as _
  31 from django.views.generic.list_detail import object_list
  32
  33 from catalogue import models
  34 from catalogue import forms
  35 from catalogue.utils import split_tags
  36 from newtagging import views as newtagging_views
  37
  38
  39 staff_required = user_passes_test(lambda user: user.is_staff)
  40
  41
  42 class LazyEncoder(simplejson.JSONEncoder):
  43     def default(self, obj):
  44         if isinstance(obj, Promise):
  45             return force_unicode(obj)
  46         return obj
  47
  48
  49 def main_page(request):
  50     if request.user.is_authenticated():
  51         shelves = models.Tag.objects.filter(category='set', user=request.user)
  52         new_set_form = forms.NewSetForm()
  53     extra_where = "NOT catalogue_tag.category = 'set'"
  54     tags = models.Tag.objects.usage_for_model(models.Book, counts=True, extra={'where': [extra_where]})
  55     fragment_tags = models.Tag.objects.usage_for_model(models.Fragment, counts=True,
  56         extra={'where': ["catalogue_tag.category = 'theme'"] + [extra_where]})
  57     categories = split_tags(tags)
  58
  59     form = forms.SearchForm()
  60     return render_to_response('catalogue/main_page.html', locals(),
  61         context_instance=RequestContext(request))
  62
  63
  64 def book_list(request):
  65     books = models.Book.objects.all()
  66     form = forms.SearchForm()
  67
  68     books_by_first_letter = SortedDict()
  69     for book in books:
  70         books_by_first_letter.setdefault(book.title[0], []).append(book)
  71
  72     return render_to_response('catalogue/book_list.html', locals(),
  73         context_instance=RequestContext(request))
  74
  75
  76 def tagged_object_list(request, tags=''):
  77     # Prevent DoS attacks on our database
  78     if len(tags.split('/')) > 6:
  79         raise Http404
  80
  81     try:
  82         tags = models.Tag.get_tag_list(tags)
  83     except models.Tag.DoesNotExist:
  84         raise Http404
  85
  86     if len([tag for tag in tags if tag.category == 'book']):
  87         raise Http404
  88
  89     theme_is_set = [tag for tag in tags if tag.category == 'theme']
  90     shelf_is_set = len(tags) == 1 and tags[0].category == 'set'
  91     my_shelf_is_set = shelf_is_set and request.user.is_authenticated() and request.user == tags[0].user
  92
  93     objects = only_author = pd_counter = None
  94     categories = {}
  95
  96     if theme_is_set:
  97         shelf_tags = [tag for tag in tags if tag.category == 'set']
  98         fragment_tags = [tag for tag in tags if tag.category != 'set']
  99         fragments = models.Fragment.tagged.with_all(fragment_tags)
 100
 101         if shelf_tags:
 102             books = models.Book.tagged.with_all(shelf_tags).order_by()
 103             l_tags = [book.book_tag() for book in books]
 104             fragments = models.Fragment.tagged.with_any(l_tags, fragments)
 105
 106         # newtagging goes crazy if we just try:
 107         #related_tags = models.Tag.objects.usage_for_queryset(fragments, counts=True,
 108         #                    extra={'where': ["catalogue_tag.category != 'book'"]})
 109         fragment_keys = [fragment.pk for fragment in fragments]
 110         if fragment_keys:
 111             related_tags = models.Fragment.tags.usage(counts=True,
 112                                 filters={'pk__in': fragment_keys},
 113                                 extra={'where': ["catalogue_tag.category != 'book'"]})
 114             related_tags = (tag for tag in related_tags if tag not in fragment_tags)
 115             categories = split_tags(related_tags)
 116
 117             objects = fragments
 118     else:
 119         # get relevant books and their tags
 120         objects = models.Book.tagged.with_all(tags).order_by()
 121         l_tags = [book.book_tag() for book in objects]
 122         # eliminate descendants
 123         descendants_keys = [book.pk for book in models.Book.tagged.with_any(l_tags)]
 124         if descendants_keys:
 125             objects = objects.exclude(pk__in=descendants_keys)
 126
 127         # get related tags from `tag_counter` and `theme_counter`
 128         related_counts = {}
 129         tags_pks = [tag.pk for tag in tags]
 130         for book in objects:
 131             for tag_pk, value in itertools.chain(book.tag_counter.iteritems(), book.theme_counter.iteritems()):
 132                 if tag_pk in tags_pks:
 133                     continue
 134                 related_counts[tag_pk] = related_counts.get(tag_pk, 0) + value
 135         related_tags = models.Tag.objects.filter(pk__in=related_counts.keys())
 136         related_tags = [tag for tag in related_tags if tag not in tags]
 137         for tag in related_tags:
 138             tag.count = related_counts[tag.pk]
 139
 140         categories = split_tags(related_tags)
 141         del related_tags
 142
 143     if not objects:
 144         only_author = len(tags) == 1 and tags[0].category == 'author'
 145         pd_counter = only_author and tags[0].goes_to_pd()
 146         objects = models.Book.objects.none()
 147
 148     return object_list(
 149         request,
 150         objects,
 151         template_name='catalogue/tagged_object_list.html',
 152         extra_context={
 153             'categories': categories,
 154             'shelf_is_set': shelf_is_set,
 155             'only_author': only_author,
 156             'pd_counter': pd_counter,
 157             'user_is_owner': my_shelf_is_set,
 158             'formats_form': forms.DownloadFormatsForm(),
 159
 160             'tags': tags,
 161         }
 162     )
 163
 164
 165 def book_fragments(request, book_slug, theme_slug):
 166     book = get_object_or_404(models.Book, slug=book_slug)
 167     book_tag = get_object_or_404(models.Tag, slug='l-' + book_slug)
 168     theme = get_object_or_404(models.Tag, slug=theme_slug)
 169     fragments = models.Fragment.tagged.with_all([book_tag, theme])
 170
 171     form = forms.SearchForm()
 172     return render_to_response('catalogue/book_fragments.html', locals(),
 173         context_instance=RequestContext(request))
 174
 175
 176 def book_detail(request, slug):
 177     try:
 178         book = models.Book.objects.get(slug=slug)
 179     except models.Book.DoesNotExist:
 180         return book_stub_detail(request, slug)
 181
 182     book_tag = book.book_tag()
 183     tags = list(book.tags.filter(~Q(category='set')))
 184     categories = split_tags(tags)
 185     book_children = book.children.all().order_by('parent_number')
 186     extra_where = "catalogue_tag.category = 'theme'"
 187     book_themes = models.Tag.objects.related_for_model(book_tag, models.Fragment, counts=True, extra={'where': [extra_where]})
 188     extra_info = book.get_extra_info_value()
 189
 190     form = forms.SearchForm()
 191     return render_to_response('catalogue/book_detail.html', locals(),
 192         context_instance=RequestContext(request))
 193
 194
 195 def book_stub_detail(request, slug):
 196     book = get_object_or_404(models.BookStub, slug=slug)
 197     pd_counter = book.pd
 198     form = forms.SearchForm()
 199
 200     return render_to_response('catalogue/book_stub_detail.html', locals(),
 201         context_instance=RequestContext(request))
 202
 203
 204 def book_text(request, slug):
 205     book = get_object_or_404(models.Book, slug=slug)
 206     book_themes = {}
 207     for fragment in book.fragments.all():
 208         for theme in fragment.tags.filter(category='theme'):
 209             book_themes.setdefault(theme, []).append(fragment)
 210
 211     book_themes = book_themes.items()
 212     book_themes.sort(key=lambda s: s[0].sort_key)
 213     return render_to_response('catalogue/book_text.html', locals(),
 214         context_instance=RequestContext(request))
 215
 216
 217 # ==========
 218 # = Search =
 219 # ==========
 220
 221 def _no_diacritics_regexp(query):
 222     """ returns a regexp for searching for a query without diacritics
 223
 224     should be locale-aware """
 225     names = {
 226         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',
 227         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'
 228         }
 229     def repl(m):
 230         l = m.group()
 231         return u"(%s)" % '|'.join(names[l])
 232     return re.sub(u'[%s]' % (u''.join(names.keys())), repl, query)
 233
 234 def unicode_re_escape(query):
 235     """ Unicode-friendly version of re.escape """
 236     return re.sub('(?u)(\W)', r'\\\1', query)
 237
 238 def _word_starts_with(name, prefix):
 239     """returns a Q object getting models having `name` contain a word
 240     starting with `prefix`
 241
 242     We define word characters as alphanumeric and underscore, like in JS.
 243
 244     Works for MySQL, PostgreSQL, Oracle.
 245     For SQLite, _sqlite* version is substituted for this.
 246     """
 247     kwargs = {}
 248
 249     prefix = _no_diacritics_regexp(unicode_re_escape(prefix))
 250     # can't use [[:<:]] (word start),
 251     # but we want both `xy` and `(xy` to catch `(xyz)`
 252     kwargs['%s__iregex' % name] = u"(^|[^[:alnum:]_])%s" % prefix
 253
 254     return Q(**kwargs)
 255
 256
 257 def _sqlite_word_starts_with(name, prefix):
 258     """ version of _word_starts_with for SQLite
 259
 260     SQLite in Django uses Python re module
 261     """
 262     kwargs = {}
 263     prefix = _no_diacritics_regexp(unicode_re_escape(prefix))
 264     kwargs['%s__iregex' % name] = ur"(^|(?<=[^\wąćęłńóśźżĄĆĘŁŃÓŚŹŻ]))%s" % prefix
 265     return Q(**kwargs)
 266
 267
 268 if settings.DATABASE_ENGINE == 'sqlite3':
 269     _word_starts_with = _sqlite_word_starts_with
 270
 271
 272 def _tags_starting_with(prefix, user=None):
 273     prefix = prefix.lower()
 274     book_stubs = models.BookStub.objects.filter(_word_starts_with('title', prefix))
 275     books = models.Book.objects.filter(_word_starts_with('title', prefix))
 276     book_stubs = filter(lambda x: x not in books, book_stubs)
 277     tags = models.Tag.objects.filter(_word_starts_with('name', prefix))
 278     if user and user.is_authenticated():
 279         tags = tags.filter(~Q(category='book') & (~Q(category='set') | Q(user=user)))
 280     else:
 281         tags = tags.filter(~Q(category='book') & ~Q(category='set'))
 282
 283     return list(books) + list(tags) + list(book_stubs)
 284
 285
 286 def _get_result_link(match, tag_list):
 287     if isinstance(match, models.Book) or isinstance(match, models.BookStub):
 288         return match.get_absolute_url()
 289     else:
 290         return reverse('catalogue.views.tagged_object_list',
 291             kwargs={'tags': '/'.join(tag.slug for tag in tag_list + [match])}
 292         )
 293
 294 def _get_result_type(match):
 295     if isinstance(match, models.Book) or isinstance(match, models.BookStub):
 296         type = 'book'
 297     else:
 298         type = match.category
 299     return dict(models.TAG_CATEGORIES)[type]
 300
 301
 302
 303 def find_best_matches(query, user=None):
 304     """ Finds a Book, Tag or Bookstub best matching a query.
 305
 306     Returns a with:
 307       - zero elements when nothing is found,
 308       - one element when a best result is found,
 309       - more then one element on multiple exact matches
 310
 311     Raises a ValueError on too short a query.
 312     """
 313
 314     query = query.lower()
 315     if len(query) < 2:
 316         raise ValueError("query must have at least two characters")
 317
 318     result = tuple(_tags_starting_with(query, user))
 319     exact_matches = tuple(res for res in result if res.name.lower() == query)
 320     if exact_matches:
 321         return exact_matches
 322     else:
 323         return result[:1]
 324
 325
 326 def search(request):
 327     tags = request.GET.get('tags', '')
 328     prefix = request.GET.get('q', '')
 329
 330     try:
 331         tag_list = models.Tag.get_tag_list(tags)
 332     except:
 333         tag_list = []
 334
 335     try:
 336         result = find_best_matches(prefix, request.user)
 337     except ValueError:
 338         return render_to_response('catalogue/search_too_short.html', {'tags':tag_list, 'prefix':prefix},
 339             context_instance=RequestContext(request))
 340
 341     if len(result) == 1:
 342         return HttpResponseRedirect(_get_result_link(result[0], tag_list))
 343     elif len(result) > 1:
 344         return render_to_response('catalogue/search_multiple_hits.html',
 345             {'tags':tag_list, 'prefix':prefix, 'results':((x, _get_result_link(x, tag_list), _get_result_type(x)) for x in result)},
 346             context_instance=RequestContext(request))
 347     else:
 348         return render_to_response('catalogue/search_no_hits.html', {'tags':tag_list, 'prefix':prefix},
 349             context_instance=RequestContext(request))
 350
 351
 352 def tags_starting_with(request):
 353     prefix = request.GET.get('q', '')
 354     # Prefix must have at least 2 characters
 355     if len(prefix) < 2:
 356         return HttpResponse('')
 357
 358     return HttpResponse('\n'.join(tag.name for tag in _tags_starting_with(prefix, request.user)))
 359
 360
 361 # ====================
 362 # = Shelf management =
 363 # ====================
 364 @login_required
 365 @cache.never_cache
 366 def user_shelves(request):
 367     shelves = models.Tag.objects.filter(category='set', user=request.user)
 368     new_set_form = forms.NewSetForm()
 369     return render_to_response('catalogue/user_shelves.html', locals(),
 370             context_instance=RequestContext(request))
 371
 372 @cache.never_cache
 373 def book_sets(request, slug):
 374     book = get_object_or_404(models.Book, slug=slug)
 375     user_sets = models.Tag.objects.filter(category='set', user=request.user)
 376     book_sets = book.tags.filter(category='set', user=request.user)
 377
 378     if not request.user.is_authenticated():
 379         return HttpResponse(_('<p>To maintain your shelves you need to be logged in.</p>'))
 380
 381     if request.method == 'POST':
 382         form = forms.ObjectSetsForm(book, request.user, request.POST)
 383         if form.is_valid():
 384             old_shelves = list(book.tags.filter(category='set'))
 385             new_shelves = [models.Tag.objects.get(pk=id) for id in form.cleaned_data['set_ids']]
 386
 387             for shelf in [shelf for shelf in old_shelves if shelf not in new_shelves]:
 388                 shelf.book_count -= 1
 389                 shelf.save()
 390
 391             for shelf in [shelf for shelf in new_shelves if shelf not in old_shelves]:
 392                 shelf.book_count += 1
 393                 shelf.save()
 394
 395             book.tags = new_shelves + list(book.tags.filter(~Q(category='set') | ~Q(user=request.user)))
 396             if request.is_ajax():
 397                 return HttpResponse(_('<p>Shelves were sucessfully saved.</p>'))
 398             else:
 399                 return HttpResponseRedirect('/')
 400     else:
 401         form = forms.ObjectSetsForm(book, request.user)
 402         new_set_form = forms.NewSetForm()
 403
 404     return render_to_response('catalogue/book_sets.html', locals(),
 405         context_instance=RequestContext(request))
 406
 407
 408 @login_required
 409 @require_POST
 410 @cache.never_cache
 411 def remove_from_shelf(request, shelf, book):
 412     book = get_object_or_404(models.Book, slug=book)
 413     shelf = get_object_or_404(models.Tag, slug=shelf, category='set', user=request.user)
 414
 415     if shelf in book.tags:
 416         models.Tag.objects.remove_tag(book, shelf)
 417
 418         shelf.book_count -= 1
 419         shelf.save()
 420
 421         return HttpResponse(_('Book was successfully removed from the shelf'))
 422     else:
 423         return HttpResponse(_('This book is not on the shelf'))
 424
 425
 426 def collect_books(books):
 427     """
 428     Returns all real books in collection.
 429     """
 430     result = []
 431     for book in books:
 432         if len(book.children.all()) == 0:
 433             result.append(book)
 434         else:
 435             result += collect_books(book.children.all())
 436     return result
 437
 438
 439 @cache.never_cache
 440 def download_shelf(request, slug):
 441     """"
 442     Create a ZIP archive on disk and transmit it in chunks of 8KB,
 443     without loading the whole file into memory. A similar approach can
 444     be used for large dynamic PDF files.
 445     """
 446     shelf = get_object_or_404(models.Tag, slug=slug, category='set')
 447
 448     formats = []
 449     form = forms.DownloadFormatsForm(request.GET)
 450     if form.is_valid():
 451         formats = form.cleaned_data['formats']
 452     if len(formats) == 0:
 453         formats = ['pdf', 'epub', 'odt', 'txt', 'mp3', 'ogg']
 454
 455     # Create a ZIP archive
 456     temp = tempfile.TemporaryFile()
 457     archive = zipfile.ZipFile(temp, 'w')
 458
 459     for book in collect_books(models.Book.tagged.with_all(shelf)):
 460         if 'pdf' in formats and book.pdf_file:
 461             filename = book.pdf_file.path
 462             archive.write(filename, str('%s.pdf' % book.slug))
 463         if 'epub' in formats and book.epub_file:
 464             filename = book.epub_file.path
 465             archive.write(filename, str('%s.epub' % book.slug))
 466         if 'odt' in formats and book.odt_file:
 467             filename = book.odt_file.path
 468             archive.write(filename, str('%s.odt' % book.slug))
 469         if 'txt' in formats and book.txt_file:
 470             filename = book.txt_file.path
 471             archive.write(filename, str('%s.txt' % book.slug))
 472         if 'mp3' in formats and book.mp3_file:
 473             filename = book.mp3_file.path
 474             archive.write(filename, str('%s.mp3' % book.slug))
 475         if 'ogg' in formats and book.ogg_file:
 476             filename = book.ogg_file.path
 477             archive.write(filename, str('%s.ogg' % book.slug))
 478     archive.close()
 479
 480     response = HttpResponse(content_type='application/zip', mimetype='application/x-zip-compressed')
 481     response['Content-Disposition'] = 'attachment; filename=%s.zip' % shelf.sort_key
 482     response['Content-Length'] = temp.tell()
 483
 484     temp.seek(0)
 485     response.write(temp.read())
 486     return response
 487
 488
 489 @cache.never_cache
 490 def shelf_book_formats(request, shelf):
 491     """"
 492     Returns a list of formats of books in shelf.
 493     """
 494     shelf = get_object_or_404(models.Tag, slug=shelf, category='set')
 495
 496     formats = {'pdf': False, 'epub': False, 'odt': False, 'txt': False, 'mp3': False, 'ogg': False}
 497
 498     for book in collect_books(models.Book.tagged.with_all(shelf)):
 499         if book.pdf_file:
 500             formats['pdf'] = True
 501         if book.epub_file:
 502             formats['epub'] = True
 503         if book.odt_file:
 504             formats['odt'] = True
 505         if book.txt_file:
 506             formats['txt'] = True
 507         if book.mp3_file:
 508             formats['mp3'] = True
 509         if book.ogg_file:
 510             formats['ogg'] = True
 511
 512     return HttpResponse(LazyEncoder().encode(formats))
 513
 514
 515 @login_required
 516 @require_POST
 517 @cache.never_cache
 518 def new_set(request):
 519     new_set_form = forms.NewSetForm(request.POST)
 520     if new_set_form.is_valid():
 521         new_set = new_set_form.save(request.user)
 522
 523         if request.is_ajax():
 524             return HttpResponse(_('<p>Shelf <strong>%s</strong> was successfully created</p>') % new_set)
 525         else:
 526             return HttpResponseRedirect('/')
 527
 528     return HttpResponseRedirect('/')
 529
 530
 531 @login_required
 532 @require_POST
 533 @cache.never_cache
 534 def delete_shelf(request, slug):
 535     user_set = get_object_or_404(models.Tag, slug=slug, category='set', user=request.user)
 536     user_set.delete()
 537
 538     if request.is_ajax():
 539         return HttpResponse(_('<p>Shelf <strong>%s</strong> was successfully removed</p>') % user_set.name)
 540     else:
 541         return HttpResponseRedirect('/')
 542
 543
 544 # ==================
 545 # = Authentication =
 546 # ==================
 547 @require_POST
 548 @cache.never_cache
 549 def login(request):
 550     form = AuthenticationForm(data=request.POST, prefix='login')
 551     if form.is_valid():
 552         auth.login(request, form.get_user())
 553         response_data = {'success': True, 'errors': {}}
 554     else:
 555         response_data = {'success': False, 'errors': form.errors}
 556     return HttpResponse(LazyEncoder(ensure_ascii=False).encode(response_data))
 557
 558
 559 @require_POST
 560 @cache.never_cache
 561 def register(request):
 562     registration_form = UserCreationForm(request.POST, prefix='registration')
 563     if registration_form.is_valid():
 564         user = registration_form.save()
 565         user = auth.authenticate(
 566             username=registration_form.cleaned_data['username'],
 567             password=registration_form.cleaned_data['password1']
 568         )
 569         auth.login(request, user)
 570         response_data = {'success': True, 'errors': {}}
 571     else:
 572         response_data = {'success': False, 'errors': registration_form.errors}
 573     return HttpResponse(LazyEncoder(ensure_ascii=False).encode(response_data))
 574
 575
 576 @cache.never_cache
 577 def logout_then_redirect(request):
 578     auth.logout(request)
 579     return HttpResponseRedirect(urlquote_plus(request.GET.get('next', '/'), safe='/?='))
 580
 581
 582
 583 # =========
 584 # = Admin =
 585 # =========
 586 @login_required
 587 @staff_required
 588 def import_book(request):
 589     """docstring for import_book"""
 590     book_import_form = forms.BookImportForm(request.POST, request.FILES)
 591     if book_import_form.is_valid():
 592         try:
 593             book_import_form.save()
 594         except:
 595             info = sys.exc_info()
 596             exception = pprint.pformat(info[1])
 597             tb = '\n'.join(traceback.format_tb(info[2]))
 598             _('Today is %(month)s, %(day)s.') % {'month': m, 'day': d}
 599             return HttpResponse(_("An error occurred: %(exception)s\n\n%(tb)s") % {'exception':exception, 'tb':tb}, mimetype='text/plain')
 600         return HttpResponse(_("Book imported successfully"))
 601     else:
 602         return HttpResponse(_("Error importing file: %r") % book_import_form.errors)
 603
 604
 605
 606 def clock(request):
 607     """ Provides server time for jquery.countdown,
 608     in a format suitable for Date.parse()
 609     """
 610     from datetime import datetime
 611     return HttpResponse(datetime.now().strftime('%Y/%m/%d %H:%M:%S'))