apps/api/handlers.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 from datetime import datetime, timedelta
   6 import json
   7 from urlparse import urljoin
   8
   9 from django.conf import settings
  10 from django.contrib.sites.models import Site
  11 from django.core.cache import get_cache
  12 from django.core.urlresolvers import reverse
  13 from piston.handler import AnonymousBaseHandler, BaseHandler
  14 from piston.utils import rc
  15
  16 from api.helpers import timestamp
  17 from api.models import Deleted
  18 from catalogue.forms import BookImportForm
  19 from catalogue.models import Book, Tag, BookMedia, Fragment
  20 from picture.models import Picture
  21 from picture.forms import PictureImportForm
  22
  23 from stats.utils import piwik_track
  24
  25 API_BASE = WL_BASE = MEDIA_BASE = 'http://' + Site.objects.get_current().domain
  26
  27
  28 category_singular = {
  29     'authors': 'author',
  30     'kinds': 'kind',
  31     'genres': 'genre',
  32     'epochs': 'epoch',
  33     'themes': 'theme',
  34     'books': 'book',
  35 }
  36 category_plural={}
  37 for k, v in category_singular.items():
  38     category_plural[v] = k
  39
  40 book_tag_categories = ['author', 'epoch', 'kind', 'genre']
  41
  42
  43
  44 def read_tags(tags, allowed):
  45     """ Reads a path of filtering tags.
  46
  47     :param str tags: a path of category and slug pairs, like: authors/an-author/...
  48     :returns: list of Tag objects
  49     :raises: ValueError when tags can't be found
  50     """
  51     if not tags:
  52         return []
  53
  54     tags = tags.strip('/').split('/')
  55     real_tags = []
  56     while tags:
  57         category = tags.pop(0)
  58         slug = tags.pop(0)
  59
  60         try:
  61             category = category_singular[category]
  62         except KeyError:
  63             raise ValueError('Unknown category.')
  64
  65         if not category in allowed:
  66             raise ValueError('Category not allowed.')
  67
  68         # !^%@#$^#!
  69         if category == 'book':
  70             slug = 'l-' + slug
  71
  72         try:
  73             real_tags.append(Tag.objects.get(category=category, slug=slug))
  74         except Tag.DoesNotExist:
  75             raise ValueError('Tag not found')
  76     return real_tags
  77
  78
  79 # RESTful handlers
  80
  81
  82 class BookMediaHandler(BaseHandler):
  83     """ Responsible for representing media in Books. """
  84
  85     model = BookMedia
  86     fields = ['name', 'type', 'url']
  87
  88     @classmethod
  89     def url(cls, media):
  90         """ Link to media on site. """
  91
  92         return MEDIA_BASE + media.file.url
  93
  94
  95 class BookDetails(object):
  96     """Custom fields used for representing Books."""
  97
  98     @classmethod
  99     def author(cls, book):
 100         return ",".join(t[0] for t in book.related_info()['tags']['author'])
 101
 102     @classmethod
 103     def href(cls, book):
 104         """ Returns an URI for a Book in the API. """
 105         return API_BASE + reverse("api_book", args=[book.slug])
 106
 107     @classmethod
 108     def url(cls, book):
 109         """ Returns Book's URL on the site. """
 110
 111         return WL_BASE + book.get_absolute_url()
 112
 113     @classmethod
 114     def children(cls, book):
 115         """ Returns all children for a book. """
 116
 117         return book.children.all()
 118
 119     @classmethod
 120     def audio(cls, book):
 121         """ Returns all media for a book. """
 122         return book.media.all()
 123
 124     @classmethod
 125     def cover(cls, book):
 126         return MEDIA_BASE + book.cover.url if book.cover else ''
 127
 128
 129
 130 class BookDetailHandler(BaseHandler, BookDetails):
 131     """ Main handler for Book objects.
 132
 133     Responsible for single Book details.
 134     """
 135     allowed_methods = ['GET']
 136     fields = ['title', 'parent', 'children'] + Book.formats + [
 137         'audio', 'url'] + book_tag_categories
 138
 139     @piwik_track
 140     def read(self, request, book):
 141         """ Returns details of a book, identified by a slug and lang. """
 142         try:
 143             return Book.objects.get(slug=book)
 144         except Book.DoesNotExist:
 145             return rc.NOT_FOUND
 146
 147
 148 class AnonymousBooksHandler(AnonymousBaseHandler, BookDetails):
 149     """ Main handler for Book objects.
 150
 151     Responsible for lists of Book objects.
 152     """
 153     allowed_methods = ('GET',)
 154     model = Book
 155     fields = ['author', 'href', 'title', 'url', 'cover']
 156
 157     @piwik_track
 158     def read(self, request, tags, top_level=False):
 159         """ Lists all books with given tags.
 160
 161         :param tags: filtering tags; should be a path of categories
 162              and slugs, i.e.: authors/an-author/epoch/an-epoch/
 163         :param top_level: if True and a book is included in the results,
 164              it's children are aren't. By default all books matching the tags
 165              are returned.
 166         """
 167         try:
 168             tags = read_tags(tags, allowed=book_tag_categories)
 169         except ValueError:
 170             return rc.NOT_FOUND
 171
 172         if tags:
 173             if top_level:
 174                 books = Book.tagged_top_level(tags)
 175                 return books if books else rc.NOT_FOUND
 176             else:
 177                 books = Book.tagged.with_all(tags)
 178         elif top_level:
 179             books = Book.objects.filter(parent=None)
 180         else:
 181             books = Book.objects.all()
 182
 183         if books.exists():
 184             return books
 185         else:
 186             return rc.NOT_FOUND
 187
 188     def create(self, request, tags, top_level=False):
 189         return rc.FORBIDDEN
 190
 191
 192 class BooksHandler(BookDetailHandler):
 193     allowed_methods = ('GET', 'POST')
 194     model = Book
 195     fields = ['author', 'href', 'title', 'url']
 196     anonymous = AnonymousBooksHandler
 197
 198     def create(self, request, tags, top_level=False):
 199         if not request.user.has_perm('catalogue.add_book'):
 200             return rc.FORBIDDEN
 201
 202         data = json.loads(request.POST.get('data'))
 203         form = BookImportForm(data)
 204         if form.is_valid():
 205             form.save()
 206             return rc.CREATED
 207         else:
 208             return rc.NOT_FOUND
 209
 210
 211 # add categorized tags fields for Book
 212 def _tags_getter(category):
 213     @classmethod
 214     def get_tags(cls, book):
 215         return book.tags.filter(category=category)
 216     return get_tags
 217 for plural, singular in category_singular.items():
 218     setattr(BookDetails, plural, _tags_getter(singular))
 219
 220 # add fields for files in Book
 221 def _file_getter(format):
 222     field = "%s_file" % format
 223     @classmethod
 224     def get_file(cls, book):
 225         f = getattr(book, field)
 226         if f:
 227             return MEDIA_BASE + f.url
 228         else:
 229             return ''
 230     return get_file
 231 for format in Book.formats:
 232     setattr(BookDetails, format, _file_getter(format))
 233
 234
 235 class TagDetails(object):
 236     """Custom Tag fields."""
 237
 238     @classmethod
 239     def href(cls, tag):
 240         """ Returns URI in the API for the tag. """
 241
 242         return API_BASE + reverse("api_tag", args=[category_plural[tag.category], tag.slug])
 243
 244     @classmethod
 245     def url(cls, tag):
 246         """ Returns URL on the site. """
 247
 248         return WL_BASE + tag.get_absolute_url()
 249
 250
 251 class TagDetailHandler(BaseHandler, TagDetails):
 252     """ Responsible for details of a single Tag object. """
 253
 254     fields = ['name', 'url', 'sort_key', 'description']
 255
 256     @piwik_track
 257     def read(self, request, category, slug):
 258         """ Returns details of a tag, identified by category and slug. """
 259
 260         try:
 261             category_sng = category_singular[category]
 262         except KeyError, e:
 263             return rc.NOT_FOUND
 264
 265         try:
 266             return Tag.objects.get(category=category_sng, slug=slug)
 267         except Tag.DoesNotExist:
 268             return rc.NOT_FOUND
 269
 270
 271 class TagsHandler(BaseHandler, TagDetails):
 272     """ Main handler for Tag objects.
 273
 274     Responsible for lists of Tag objects
 275     and fields used for representing Tags.
 276
 277     """
 278     allowed_methods = ('GET',)
 279     model = Tag
 280     fields = ['name', 'href', 'url']
 281
 282     @piwik_track
 283     def read(self, request, category):
 284         """ Lists all tags in the category (eg. all themes). """
 285
 286         try:
 287             category_sng = category_singular[category]
 288         except KeyError, e:
 289             return rc.NOT_FOUND
 290
 291         tags = Tag.objects.filter(category=category_sng).exclude(book_count=0)
 292         if tags.exists():
 293             return tags
 294         else:
 295             return rc.NOT_FOUND
 296
 297
 298 class FragmentDetails(object):
 299     """Custom Fragment fields."""
 300
 301     @classmethod
 302     def href(cls, fragment):
 303         """ Returns URI in the API for the fragment. """
 304
 305         return API_BASE + reverse("api_fragment",
 306             args=[fragment.book.slug, fragment.anchor])
 307
 308     @classmethod
 309     def url(cls, fragment):
 310         """ Returns URL on the site for the fragment. """
 311
 312         return WL_BASE + fragment.get_absolute_url()
 313
 314     @classmethod
 315     def themes(cls, fragment):
 316         """ Returns a list of theme tags for the fragment. """
 317
 318         return fragment.tags.filter(category='theme')
 319
 320
 321 class FragmentDetailHandler(BaseHandler, FragmentDetails):
 322     fields = ['book', 'anchor', 'text', 'url', 'themes']
 323
 324     @piwik_track
 325     def read(self, request, book, anchor):
 326         """ Returns details of a fragment, identified by book slug and anchor. """
 327         try:
 328             return Fragment.objects.get(book__slug=book, anchor=anchor)
 329         except Fragment.DoesNotExist:
 330             return rc.NOT_FOUND
 331
 332
 333 class FragmentsHandler(BaseHandler, FragmentDetails):
 334     """ Main handler for Fragments.
 335
 336     Responsible for lists of Fragment objects
 337     and fields used for representing Fragments.
 338
 339     """
 340     model = Fragment
 341     fields = ['book', 'url', 'anchor', 'href']
 342     allowed_methods = ('GET',)
 343
 344     categories = set(['author', 'epoch', 'kind', 'genre', 'book', 'theme'])
 345
 346     @piwik_track
 347     def read(self, request, tags):
 348         """ Lists all fragments with given book, tags, themes.
 349
 350         :param tags: should be a path of categories and slugs, i.e.:
 351              books/book-slug/authors/an-author/themes/a-theme/
 352
 353         """
 354         try:
 355             tags = read_tags(tags, allowed=self.categories)
 356         except ValueError:
 357             return rc.NOT_FOUND
 358         fragments = Fragment.tagged.with_all(tags).select_related('book')
 359         if fragments.exists():
 360             return fragments
 361         else:
 362             return rc.NOT_FOUND
 363
 364
 365
 366 # Changes handlers
 367
 368 class CatalogueHandler(BaseHandler):
 369
 370     @staticmethod
 371     def fields(request, name):
 372         fields_str = request.GET.get(name) if request is not None else None
 373         return fields_str.split(',') if fields_str is not None else None
 374
 375     @staticmethod
 376     def until(t=None):
 377         """ Returns time suitable for use as upper time boundary for check.
 378
 379             Used to avoid issues with time between setting the change stamp
 380             and actually saving the model in database.
 381             Cuts the microsecond part to avoid issues with DBs where time has
 382             more precision.
 383
 384             :param datetime t: manually sets the upper boundary
 385
 386         """
 387         # set to five minutes ago, to avoid concurrency issues
 388         if t is None:
 389             t = datetime.now() - timedelta(seconds=settings.API_WAIT)
 390         # set to whole second in case DB supports something smaller
 391         return t.replace(microsecond=0)
 392
 393     @staticmethod
 394     def book_dict(book, fields=None):
 395         all_fields = ['url', 'title', 'description',
 396                       'gazeta_link', 'wiki_link',
 397                       ] + Book.formats + BookMedia.formats + [
 398                       'parent', 'parent_number',
 399                       'tags',
 400                       'license', 'license_description', 'source_name',
 401                       'technical_editors', 'editors',
 402                       'author', 'sort_key',
 403                      ]
 404         if fields:
 405             fields = (f for f in fields if f in all_fields)
 406         else:
 407             fields = all_fields
 408
 409         extra_info = book.get_extra_info_value()
 410
 411         obj = {}
 412         for field in fields:
 413
 414             if field in Book.formats:
 415                 f = getattr(book, field+'_file')
 416                 if f:
 417                     obj[field] = {
 418                         'url': f.url,
 419                         'size': f.size,
 420                     }
 421
 422             elif field in BookMedia.formats:
 423                 media = []
 424                 for m in book.media.filter(type=field):
 425                     media.append({
 426                         'url': m.file.url,
 427                         'size': m.file.size,
 428                     })
 429                 if media:
 430                     obj[field] = media
 431
 432             elif field == 'url':
 433                 obj[field] = book.get_absolute_url()
 434
 435             elif field == 'tags':
 436                 obj[field] = [t.id for t in book.tags.exclude(category__in=('book', 'set'))]
 437
 438             elif field == 'author':
 439                 obj[field] = ", ".join(t.name for t in book.tags.filter(category='author'))
 440
 441             elif field == 'parent':
 442                 obj[field] = book.parent_id
 443
 444             elif field in ('license', 'license_description', 'source_name',
 445                       'technical_editors', 'editors'):
 446                 f = extra_info.get(field)
 447                 if f:
 448                     obj[field] = f
 449
 450             else:
 451                 f = getattr(book, field)
 452                 if f:
 453                     obj[field] = f
 454
 455         obj['id'] = book.id
 456         return obj
 457
 458     @classmethod
 459     def book_changes(cls, request=None, since=0, until=None, fields=None):
 460         since = datetime.fromtimestamp(int(since))
 461         until = cls.until(until)
 462
 463         changes = {
 464             'time_checked': timestamp(until)
 465         }
 466
 467         if not fields:
 468             fields = cls.fields(request, 'book_fields')
 469
 470         added = []
 471         updated = []
 472         deleted = []
 473
 474         last_change = since
 475         for book in Book.objects.filter(changed_at__gte=since,
 476                     changed_at__lt=until):
 477             book_d = cls.book_dict(book, fields)
 478             updated.append(book_d)
 479         if updated:
 480             changes['updated'] = updated
 481
 482         for book in Deleted.objects.filter(content_type=Book,
 483                     deleted_at__gte=since,
 484                     deleted_at__lt=until,
 485                     created_at__lt=since):
 486             deleted.append(book.id)
 487         if deleted:
 488             changes['deleted'] = deleted
 489
 490         return changes
 491
 492     @staticmethod
 493     def tag_dict(tag, fields=None):
 494         all_fields = ('name', 'category', 'sort_key', 'description',
 495                       'gazeta_link', 'wiki_link',
 496                       'url', 'books',
 497                      )
 498
 499         if fields:
 500             fields = (f for f in fields if f in all_fields)
 501         else:
 502             fields = all_fields
 503
 504         obj = {}
 505         for field in fields:
 506
 507             if field == 'url':
 508                 obj[field] = tag.get_absolute_url()
 509
 510             elif field == 'books':
 511                 obj[field] = [b.id for b in Book.tagged_top_level([tag])]
 512
 513             elif field == 'sort_key':
 514                 obj[field] = tag.sort_key
 515
 516             else:
 517                 f = getattr(tag, field)
 518                 if f:
 519                     obj[field] = f
 520
 521         obj['id'] = tag.id
 522         return obj
 523
 524     @classmethod
 525     def tag_changes(cls, request=None, since=0, until=None, fields=None, categories=None):
 526         since = datetime.fromtimestamp(int(since))
 527         until = cls.until(until)
 528
 529         changes = {
 530             'time_checked': timestamp(until)
 531         }
 532
 533         if not fields:
 534             fields = cls.fields(request, 'tag_fields')
 535         if not categories:
 536             categories = cls.fields(request, 'tag_categories')
 537
 538         all_categories = ('author', 'epoch', 'kind', 'genre')
 539         if categories:
 540             categories = (c for c in categories if c in all_categories)
 541         else:
 542             categories = all_categories
 543
 544         updated = []
 545         deleted = []
 546
 547         for tag in Tag.objects.filter(category__in=categories,
 548                     changed_at__gte=since,
 549                     changed_at__lt=until):
 550             # only serve non-empty tags
 551             if tag.book_count:
 552                 tag_d = cls.tag_dict(tag, fields)
 553                 updated.append(tag_d)
 554             elif tag.created_at < since:
 555                 deleted.append(tag.id)
 556         if updated:
 557             changes['updated'] = updated
 558
 559         for tag in Deleted.objects.filter(category__in=categories,
 560                 content_type=Tag,
 561                     deleted_at__gte=since,
 562                     deleted_at__lt=until,
 563                     created_at__lt=since):
 564             deleted.append(tag.id)
 565         if deleted:
 566             changes['deleted'] = deleted
 567
 568         return changes
 569
 570     @classmethod
 571     def changes(cls, request=None, since=0, until=None, book_fields=None,
 572                 tag_fields=None, tag_categories=None):
 573         until = cls.until(until)
 574         since = int(since)
 575
 576         if not since:
 577             cache = get_cache('api')
 578             key = hash((book_fields, tag_fields, tag_categories,
 579                     tuple(sorted(request.GET.items()))
 580                   ))
 581             value = cache.get(key)
 582             if value is not None:
 583                 return value
 584
 585         changes = {
 586             'time_checked': timestamp(until)
 587         }
 588
 589         changes_by_type = {
 590             'books': cls.book_changes(request, since, until, book_fields),
 591             'tags': cls.tag_changes(request, since, until, tag_fields, tag_categories),
 592         }
 593
 594         for model in changes_by_type:
 595             for field in changes_by_type[model]:
 596                 if field == 'time_checked':
 597                     continue
 598                 changes.setdefault(field, {})[model] = changes_by_type[model][field]
 599
 600         if not since:
 601             cache.set(key, changes)
 602
 603         return changes
 604
 605
 606 class BookChangesHandler(CatalogueHandler):
 607     allowed_methods = ('GET',)
 608
 609     @piwik_track
 610     def read(self, request, since):
 611         return self.book_changes(request, since)
 612
 613
 614 class TagChangesHandler(CatalogueHandler):
 615     allowed_methods = ('GET',)
 616
 617     @piwik_track
 618     def read(self, request, since):
 619         return self.tag_changes(request, since)
 620
 621
 622 class ChangesHandler(CatalogueHandler):
 623     allowed_methods = ('GET',)
 624
 625     @piwik_track
 626     def read(self, request, since):
 627         return self.changes(request, since)
 628
 629
 630 class PictureHandler(BaseHandler):
 631     model = Picture
 632     fields = ('slug', 'title')
 633     allowed_methods = ('POST',)
 634
 635     def create(self, request):
 636         if not request.user.has_perm('picture.add_picture'):
 637             return rc.FORBIDDEN
 638
 639         data = json.loads(request.POST.get('data'))
 640         form = PictureImportForm(data)
 641         if form.is_valid():
 642             form.save()
 643             return rc.CREATED
 644         else:
 645             return rc.NOT_FOUND