The previous implementation only supported rendering
attachments within the context of a node or project document.
Now it also supports node.properties. This is a temporary
solution, as noted in the TODO comments.
During the years we went from site-wide blog, to project blog, to
post view inside a project, to full one-page post view. This led
to have multiple ways to see the same content.
This commit brings all post related stuff to always use index.pug
(or index_archive if we are looking blasts from the past).
This function generates a list of selected links for important nodes such
as Pages and Blog. This list of links is used in the templates to provide
high level navigation of a Project.
Synchronisation is performed in the background by the Celery Beat, every
10 minutes. It has a time limit of 9 minutes to prevent multiple refresh
tasks from running at the same time.
Synchronisation is also possible with the `manage.py badges sync` CLI
command, which can sync either a single user or all users.
This also changes the way we treat Blender ID tokens. Before, the Blender ID
token was discarded and a random token was generated & stored. Now the
actual Blender ID token is stored.
The Facebook and Google OAuth code still uses the old approach of generating
a new token. Not sure what the added value is, though, because once the
Django session is gone there is nothing left to authenticate the user and
thus the random token is useless anyway.
This partially reverts commit c57aefd48b.
The code to check against hashed tokens remains, because existing tokens
should still work.
The unhashed tokens are necessary for fetching badges from Blender ID.
Video progress updates:
- Mark as 'done' when 90% or more is watched.
- Keep 'done' flag when re-watching.
The video progress is stored on three events, whichever comes first:
- Every 30 seconds of video.
- Every 10% of the video.
- Every pause/stop/navigation to another page.
- When we detect the video is looping.
Apparently Eve is now stricter in checking against MONGO_QUERY_BLACKLIST,
and blocks our use of $regex when getting child nodes. See
`jstree.py::jstree_get_children()`
String concatenation is bound to mess up; in this case it was producing
double slashes instead of single ones when `BLENDER_ID_ENDPOINT` ends in
a slash. Since URLs generally end in a slash, this should be supported.
We now support both the old coerce=markdown and the new validator=markdown.
Probably support for the old can be removed, but I'm keeping it around
just to be sure.
The post_internal function does `document = validator.document`, replacing
the to-be-posted document by the copy that Cerberus made (and which we
cannot add keys to because it iterates over the keys and the dict size thus
isn't allowed to change).
I hope this doesn't break other validators who expect to be able to write
to `self.document`.
Eve's Validator has not only a validate() function, but also
validate_update() and validate_replace(). Those set
self.persisted_document, so if that attribute exists we just use it.
Custom types became rather useless in Cerberus 1.0 since the type checker
is cripled (doesn't know field name, cannot return useful/detailed error
messages). Instead we use a validator now.
