Set the sender name and IP addresses explicitly in the controller rather
than implicitly in the model.
Fixes cases where automated dmails from DanbooruBot had their IP
addresses set to the person who triggered the dmail, even though they
didn't actually send the dmail themselves.
Make searches on the /tags index includes aliases too. Show matching
aliases like this:
Name: gray*
? 75098 grey_hair <- gray_hair
? 35345 grey_eyes <- gray_eyes
Refactor to use a recursive CTE to calculate implied tags in SQL, rather
than storing them in a descendant_names field. This avoids the
complexity of keeping the stored field up to date. It's also more
flexible, since it allows us to find both descendant tags (tags that
imply a given tag) as well as ancestor tags (tags that are implied by a
given tag).
Remove the soft requirement that the destination tag must have 50 posts
in order to request an alias. This often gets in the way when trying to
rename a tag to a new name that doesn't already exist.
Remove the edit, update, and approve endpoints for tag aliases and
implications. These have been useless since individual alias and
implication requests were removed. Aliases and implications could only
be edited or approved if they were in the pending state, which is no
longer possible.
Also remove unused new alias/implication request forms.
Remove the approve option from the post mode menu. Mass approving posts
is rarely needed. The javascript was also buggy (it didn't update the
data attributes correctly when undeleting a post).
The replacement for this feature is to use a tagscript with the
status:active tag.
* Allow approvers to approve a post by tagging it with status:active.
* Allow approvers to ban a post by tagging it with status:banned.
* Allow approvers to unban a post by tagging it with -status:banned.
Remove `POST /moderator/post/undelete` endpoint. Replace it with
`POST /post_approvals` instead.
Fixes it so that undeleting a post has the same behavior as approving a
post. Namely, it reloads the page instead of just flashing a "Post was
undeleted" message.
Refactor how model visibility works in index actions:
* Call `visible` in the controller instead of in model `search`
methods. This decouples model visibility from model searching.
* Explicitly pass CurrentUser when calling `visible`. This reduces
hidden dependencies on the current user inside models.
* Standardize on calling the method `visible`. In some places it was
called `permitted` instead.
* Add a `visible` base method to ApplicationModel.
Standardize on using `show_deleted: true` instead of `tags: "status:any"`
when rendering thumbnails and we want to show deleted thumbnails.
Also fixes it so that deleted thumbnails are shown when reordering pools
and favorite groups.
Remove options for searching for "Artist requested removal" and
"duplicate" flag reasons. These were legacy flag reasons that haven't
been used for a long time.
Remove the creator_id field from artists, notes, and pools. The
creator_id wasn't otherwise used and was inconsistent with the
artist/note/pool history in some cases, especially for old artists.
* Don't show who disapproved the post to other users. Only show the
creator to mods or to the disapprover themselves.
* Let unprivileged users see the /post_disapprovals index.
Remove the post update count estimate from BUR show pages. This was
complex, slow, and usually inaccurate since it assumed that requests in
a BUR had no overlap with each other, which usually wasn't the case.
Format scripts on the BUR index page as dtext, the same way they're
formatted in forum posts. This means that tags have colors and underlines,
but they don't have post counts. Gathering post counts was too slow.
Refactor to use accepts_nested_attributes_for instead of the notes
attribute to facilitate editing wikis on the artist edit page.
This fixes the notes attribute unintentionally showing up in the API.
This also changes it so that renaming an artist entry doesn't
automatically rename the corresponding wiki page. This had bad behavior
when there was a conflict between wiki pages (the wikis would be
silently merged, which usually isn't what you want). It also didn't warn
about wiki links being broken by renames.
* Replace the .category-N CSS classes on tags with .tag-type-N. Before
we were inconsistent about whether tag colors were indicated with
.category-N or .tag-type-N. Now it's always .tag-type-N.
* Fix various places to not use Tag.category_for. Tag.category_for does
one Redis call per tag lookup, which leads to N Redis calls on many
pages. This was inefficient because usually we either already had the
tags from the database, or we could fetch them easily.
Don't make /dmails default to the received folder when no other folder
is specified. This is surprising in the API because it means
/dmails.json only shows received dmails by default, not all dmails. This
isn't necessary either since everything that links to dmails already
specifies the received folder.
Remove various associated fields that were included by default on
certain endpoints. API users can use the only param to include the
full association if they need these fields.
* /artists.json: urls.
* /artist_urls.json: artist.
* /comments.json: creator_name and updater_name.
* /notes.json: creator_name.
* /pools.json: creator_name.
* /posts.json: uploader_name, children_ids, pixiv_ugoira_frame_data.
* /post_appeals.json: is_resolved.
* /post_versions.json: updater_name.
* /uploads.json: uploader_name.
Remove the list of most-used source domains from artist summaries. This
took up a lot of space and usually wasn't very useful. It was also slow.
We had to calculate this on every artist tag search so we could display
it in the Artist tab, even though usually the user didn't open the tab.
* Link other names to artist searches instead of to the new artist page.
* Remove the asterisk next to other names indicating the name isn't used
as the primary name of another artist. This is almost always the case.
Fix exception when trying to get the category_name for the html data
attributes. The wiki title is nil here and Tag.category_for couldn't
handle a nil tag name.
* Enable word breaks for all DText-formatted text. Ensures extremely
long strings in comments get broken.
* Fix the author column on comments/forum posts to use a fixed 12em
width (except on the mobile layout). Using flex-basis was wrong because
it allowed the author column to shrink if the other column was too wide.
Fix an invalid SQL exception that occurs when the paginator tries to
do a COUNT(*) to calculate the page count of a relation that already
includes a GROUP BY + COUNT(*) clause. We need to nest the whole query
inside a `SELECT COUNT(*) FROM (...)` subquery so the inner COUNT(*)
doesn't mess up the outer COUNT(*).
Fixes#4285.
Calling next_page here raises an exception when calculating the page
count because it can't handle SQL containing a GROUP BY clause. Swallow
the exception as a temp fix.
Fix an exception when the user's unread_dmail_count is nonzero but they
don't have any unread dmails. This normally shouldn't happen unless the
unread_dmail_count is inconsistent.
* Automatically generate a mod report when a comment, forum post, or
dmail is detected as spam.
* Automatically ban users that receive too many automatic spam reports
within a short window of time.
* Automatically mark spam dmails as deleted.
* Change ban threshold from 10 spam reports in 24 hours to 10 reports in 1 hour.
* Change ban length from 3 days to forever.
Fix a couple security issues related to dmail permalinks. Dmails have a
permalink that you can give to a Mod to let them read the dmail. This is
done with a key param that grants access when the dmail is opened by
another user. The key param had several problems:
* The key contained a full copy of the message's title and body encoded in
base64. This meant that anyone given a dmail permalink could read the
full dmail just by decoding the key in the link, without even having
to open the link.
* The key was derived from the dmail's title and body. If you knew or
could guess a dmail's title and body you could open the dmail. One
case when this was possible was when sending dmails. You could send
someone a dmail, take the permalink from your sent copy of the dmail,
then increment the dmail id to open the receiver's copy of the dmail.
Since the sent copy and the received copy both had the same title and
body, they both had the same dmail key. This let you check whether a
person had read your dmail, and what time they read it at.
* The key verification was done with an insecure string comparison
rather than a secure constant-time comparison. This was potentially
vulnerable to timing attacks.
* Opening a dmail belonging to another user would mark it as read for them.
The fix to all this is to use the dmail's id as the key instead of the
dmail's title and body. This means that old permalinks no longer work.
This is unavoidable given the issues above.
Other changes:
* The name of the 'Permalink' link is now 'Share'.
* Anyone with the 'Share' link can view the dmail, not just Mods.
evazion
evazion