# GNU MediaGoblin -- federated, autonomous media hosting # Copyright (C) 2011, 2012 MediaGoblin contributors. See AUTHORS. # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published by # the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see . """ This module contains some Mixin classes for the db objects. A bunch of functions on the db objects are really more like "utility functions": They could live outside the classes and be called "by hand" passing the appropiate reference. They usually only use the public API of the object and rarely use database related stuff. These functions now live here and get "mixed in" into the real objects. """ from mediagoblin import mg_globals from mediagoblin.auth import lib as auth_lib from mediagoblin.media_types import get_media_manager from mediagoblin.tools import common, licenses from mediagoblin.tools.text import cleaned_markdown_conversion from mediagoblin.tools.url import slugify class UserMixin(object): def check_login(self, password): """ See if a user can login with this password """ return auth_lib.bcrypt_check_password( password, self.pw_hash) @property def bio_html(self): return cleaned_markdown_conversion(self.bio) class MediaEntryMixin(object): def generate_slug(self): # import this here due to a cyclic import issue # (db.models -> db.mixin -> db.util -> db.models) from mediagoblin.db.util import check_media_slug_used self.slug = slugify(self.title) duplicate = check_media_slug_used(mg_globals.database, self.uploader, self.slug, self.id) if duplicate: if self.id is not None: self.slug = u"%s-%s" % (self.id, self.slug) else: self.slug = None @property def description_html(self): """ Rendered version of the description, run through Markdown and cleaned with our cleaning tool. """ return cleaned_markdown_conversion(self.description) def get_display_media(self, media_map, fetch_order=common.DISPLAY_IMAGE_FETCHING_ORDER): """ Find the best media for display. Args: - media_map: a dict like {u'image_size': [u'dir1', u'dir2', u'image.jpg']} - fetch_order: the order we should try fetching images in Returns: (media_size, media_path) """ media_sizes = media_map.keys() for media_size in common.DISPLAY_IMAGE_FETCHING_ORDER: if media_size in media_sizes: return media_map[media_size] def main_mediafile(self): pass @property def slug_or_id(self): return (self.slug or self._id) def url_for_self(self, urlgen, **extra_args): """ Generate an appropriate url for ourselves Use a slug if we have one, else use our '_id'. """ uploader = self.get_uploader return urlgen( 'mediagoblin.user_pages.media_home', user=uploader.username, media=self.slug_or_id, **extra_args) @property def thumb_url(self): """Return the thumbnail URL (for usage in templates) Will return either the real thumbnail or a default fallback icon.""" # TODO: implement generic fallback in case MEDIA_MANAGER does # not specify one? if u'thumb' in self.media_files: thumb_url = mg_globals.app.public_store.file_url( self.media_files[u'thumb']) else: # No thumbnail in media available. Get the media's # MEDIA_MANAGER for the fallback icon and return static URL # Raise FileTypeNotSupported in case no such manager is enabled manager = get_media_manager(self.media_type) thumb_url = mg_globals.app.staticdirector(manager[u'default_thumb']) return thumb_url def get_fail_exception(self): """ Get the exception that's appropriate for this error """ if self.fail_error: return common.import_component(self.fail_error) def get_license_data(self): """Return license dict for requested license""" return licenses.SUPPORTED_LICENSES[self.license or ""] def exif_display_iter(self): from mediagoblin.tools.exif import USEFUL_TAGS if not self.media_data: return exif_all = self.media_data.get("exif_all") for key in USEFUL_TAGS: if key in exif_all: yield key, exif_all[key] class MediaCommentMixin(object): @property def content_html(self): """ the actual html-rendered version of the comment displayed. Run through Markdown and the HTML cleaner. """ return cleaned_markdown_conversion(self.content) class CollectionMixin(object): def generate_slug(self): # import this here due to a cyclic import issue # (db.models -> db.mixin -> db.util -> db.models) from mediagoblin.db.util import check_collection_slug_used self.slug = slugify(self.title) duplicate = check_collection_slug_used(mg_globals.database, self.creator, self.slug, self.id) if duplicate: if self.id is not None: self.slug = u"%s-%s" % (self.id, self.slug) else: self.slug = None @property def description_html(self): """ Rendered version of the description, run through Markdown and cleaned with our cleaning tool. """ return cleaned_markdown_conversion(self.description) @property def slug_or_id(self): return (self.slug or self._id) def url_for_self(self, urlgen, **extra_args): """ Generate an appropriate url for ourselves Use a slug if we have one, else use our '_id'. """ creator = self.get_creator return urlgen( 'mediagoblin.user_pages.user_collection', user=creator.username, collection=self.slug_or_id, **extra_args) class CollectionItemMixin(object): @property def note_html(self): """ the actual html-rendered version of the note displayed. Run through Markdown and the HTML cleaner. """ return cleaned_markdown_conversion(self.note)