Specify the default settings for new users inside the User model instead
of inside the database. This makes it easier to change defaults, and it
makes the code clearer.
Remove the enable_post_navigation option. This option was originally
added to disable the next/prev post navbar beneath posts. It was later
repurposed to disable keyboard shortcuts.
Users who don't want keyboard shortcuts are advised to not press random
buttons on the keyboard like a caveman.
Only ~1200 users disabled this option and only ~600 were seen in the
last year.
Remove the enable_sequential_post_navigation option. This option was
used to disable the next/previous post navbar below posts.
This option was originally added in issue #674 because of people
complaining about the navbar when it was originally added. Also there
were complaints about URLs being uglier because of search params in the
URL (e.g. /posts/1234?q=touhou). There were also various minor bugs with
it at the time, such as keyboard shortcuts not working correctly, or the
page not remembering your search after a tag edit.
These complaints are irrelevant nowadays because a) people are used to
the navbar by now (and more often complain about it *not* being there
for order:score searches), b) post URLs always contain the search now,
this option hasn't disabled that for years, and c) the initial bugs with
it were fixed years ago.
Only ~1000 users disabled this option and only ~600 were seen in the last year.
Users still wishing to hide the search navbar can use custom CSS instead.
Include the "q" URL param (as in `/posts/1234?q=<search>`) on thumbnails
for all users. Previously it was only for logged in users. This lets the
next/previous post navbar beneath posts be used by logged out users.
Refactor the post preview html to use the ViewComponent framework. This
lets us encapsulate all the HTML, CSS, and helper methods for a UI
component in a single place.
See https://viewcomponent.org.
Adjust permissions on user events to let Moderators only see login,
logout, and user creation events, not other types of events (password
changes, etc). Admins can see everything. These other types of events
are meant for account security purposes and aren't very relevant for
sockpuppet detection purposes.
When a user does a tag search, log a few more things, including the normalized
search string, the number of tags in the search string, and the number of results.
Refactor page limits to a) be explicitly listed in the User class (not
hidden away in the Danbooru config) and b) explicitly depend on the
CurrentUser (not implicitly by way of Danbooru.config.max_numbered_pages).
Optimize autocomplete to ignore various types of bogus input that will
never match anything. It turns out it's not uncommon for people to do
things like paste random URLs into autocomplete, or hold down keys, or
enter long strings of gibberish text (sometimes in other languages).
Some things, like autocorrect and slash abbreviations, become
pathologically slow when fed certain types of bad input.
Autocomplete will abort and return nothing in the following situations:
* Searching for URLs (tags that start with http:// or https://).
* Overly long tags (strings longer than the 170 char tag name limit).
* Slash abbreviations longer than 10 chars (e.g. typing `/qwoijqoiqogirqewgoi`).
* Slash abbreviations that aren't alphanumeric (e.g. typing `/////////`).
* Autocorrect input that contains too much punctuation and not enough actual letters.
Optimize searches for non-English phrases in autocomplete. These
searches were pretty slow, and could sometimes cause sitewide lag spikes
when users typed long strings of non-English text into the search box
and caused an unintentional DoS.
The trick is to use an `array_to_tsvector(other_names) USING gin` index
on other_names. This supports fast string prefix matching against all
elements of the array. The downside is that it doesn't allow infix or
suffix matches, so we can't support wildcards in general. Wildcards
didn't quite work anyway, since artist and wiki other names can contain
literal '*' characters.
Fix `normalize_whitespace` to not strip zero-width joiner characters
(U+200D). These characters are used in emoji and stripping them breaks
some artist other names that use emoji.
Fix the `normalize` and `array_attribute` macros conflicting with each
other on the WikiPage model. This meant code like
`wiki_page.other_names = "foo bar"` didn't work. Both macros defined a
`other_names=` method, but one method overrode the other.
The fix is to use anonymous modules and prepend so we can chain method
calls with super.
Fix wiki pages and artists to normalize other names more consistently
and correctly.
For wiki pages, we strip leading / trailing / repeated underscores to
fix user typos, and we normalize to NFKC form to make search more consistent.
For artists, we allow leading / trailing / repeated underscores because
some artist names have these, and we normalize to NFC form because some
artists have weird names that would be lost by NFKC form. This does make
search less consistent.
Fix saved searches to remove additional invalid characters from labels:
* Remove repeated spaces or underscores.
* Remove leading and trailing spaces or underscores.
* Normalize Unicode characters to NFC form.
Also add a fix script to renormalize labels in old saved searches. A few
problems with existing searches:
* Some saved searches somehow had labels containing NULL elements.
* Some had leading or trailing underscores.
* Some had repeated underscores.
* Some had non-English characters in uppercase.
Add a custom Shoulda matcher for testing that a model correctly normalizes an attribute.
Usage:
subject { build(:wiki_page) }
should normalize_attribute(:title).from(" Azur Lane ").to("azur_lane")
Add tracking of certain important user actions. These events include:
* Logins
* Logouts
* Failed login attempts
* Account creations
* Account deletions
* Password reset requests
* Password changes
* Email address changes
This is similar to the mod actions log, except for account activity
related to a single user.
The information tracked includes the user, the event type (login,
logout, etc), the timestamp, the user's IP address, IP geolocation
information, the user's browser user agent, and the user's session ID
from their session cookie. This information is visible to mods only.
This is done with three models. The UserEvent model tracks the event
type (login, logout, password change, etc) and the user. The UserEvent
is tied to a UserSession, which contains the user's IP address and
browser metadata. Finally, the IpGeolocation model contains the
geolocation information for IPs, including the city, country, ISP, and
whether the IP is a proxy.
This tracking will be used for a few purposes:
* Letting users view their account history, to detect things like logins
from unrecognized IPs, failed logins attempts, password changes, etc.
* Rate limiting failed login attempts.
* Detecting sockpuppet accounts using their login history.
* Detecting unauthorized account sharing.
Add a Restricted user level. Restricted users are level 10, below
Members. New users start out as Restricted if they sign up from a proxy
or an IP recently used by another user.
Restricted users can't update or edit any public content on the site
until they verify their email address, at which point they're promoted
to Member. Restricted users are only allowed to do personal actions
like keep favorites, keep favgroups and saved searches, mark dmails as
read or deleted, or mark forum posts as read.
The restricted state already existed before, the only change here is
that now it's an actual user level instead of a hidden state. Before it
was based on two hidden flags on the user, the `requires_verification`
flag (set when a user signs up from a proxy, etc), and the `is_verified`
flag (set after the user verifies their email). Making it a user level
means that now the Restricted status will be shown publicly.
Introducing a new level below Member means that we have to change every
`is_member?` check to `!is_anonymous` for every place where we used
`is_member?` to check that the current user is logged in.
Remove the WikiPageVersion#artist and Post#updater associations. Neither
of these existed and they caused problems with searching includable
associations in the API.
In Danbooru 1, aliases (and implications) had a `reason` field where
either the admin or the alias requester gave a reason for the alias.
This field was removed from the code and the database schema, but it
still existed in the production database. This adds the field back, so
that the dev schema is consistent with the production schema, and so
that legacy reasons can be viewed on site again.
* Add back legacy tag_aliases.reason and tag_implications.reason field.
* Make /tag_aliases and /tag_implications show legacy reasons.
* Add the reason field to the search form.
Fixes#1862. Very old posts with transparent backgrounds used black
instead of white backgrounds in thumbnails. This was changed in #603
(see also #1239), but the thumbnails were never regenerated.
Purge cached thumbnails from Cloudflare when a post is regenerated.
This is necessary because regenerating a post may change the thumbnail,
and if we don't purge the cache from Cloudflare then users will still
see the old thumbnail.
We have do this before updating IQDB because if we don't, IQDB will see
the old cached thumbnail and index the wrong image. This may be racy
because the thumbnail might not be completely purged from Cloudflare
before it's downloaded by IQDB.
Regenerate posts asynchronously using a delayed job.
Regenerating a post can be slow because it involves downloading the
original file, regenerating the thumbnails, and redistributing the new
thumbnails back to the image servers. It's better to run this in the
background, especially if a user is trying to regenerate posts in bulk.
The downside is there's no notification to the user when the regeneration
is complete. You have to check the modactions log to see when it's finished.
* Remove the PostRegeneration model. Instead just use a mod action
to log when a post is regenerated.
* Change it so that IQDB is also updated when the image samples are
regenerated. This is necessary because when the images samples are
regenerated, the thumbnail may change, which means IQDB needs to be
updated too. This can happen when regenerating old images with
transparent backgrounds where the transparency was flattened to black
instead of white in the thumbnail.
* Only display one "Regenerate image" option in the post sidebar, to
regenerate both the images and IQDB. Regenerating IQDB only can be
done through the API. Having two options in the sidebar is too much
clutter, and it's too confusing for Mods who don't know the difference
between an IQDB-only regeneration and a full image regeneration.
* Add a confirm prompt to the "Regenerate image" link.
Disable the browser's native spellchecking ability on all form inputs,
except for DText inputs. We do this by setting `spellcheck="false"` on
the <body> tag, and `spellcheck="true"` on DText <input> tags.
This fixes browsers displaying a red wavy underline beneath tags in the
tag search box, among other places. We disable spellchecking globally
because most form inputs, except for DText inputs, aren't meant for
natural English language.
Display a red wavy underline beneath misspelled tags in autocomplete.
We use an inline image for the underline instead of the native
`text-decoration: red wavy underline` property because the native
underline is too big and ugly, and we have no way to adjust it. Making a
nice-looking wavy underline in CSS is surprisingly difficult. This
turned out to be the cleanest way.
* Remove the data-is-favorited attribute from post thumbnails.
* Remove the is_favorited attribute from the /posts.json API.
* Remove the fav_string attribute from the /posts.json API (only visible
to moderators).
* Change `Post#favorited_by?` to not use the fav_string.
Further addresses #4652 by eliminating the last places where fav_string
was used.
Refactor fav:<name> and ordfav:<name> searches to use the favorites
table instead of the posts.fav_string.
This may be slower for fav:<name> searches. The fav_string effectively
treats favorites like secret tags on the post, so fav:<name> searches
were effectively the same as tag searches. Now they do a subquery on the
favorites table, which may not perform as well for things like multiple
fav:<name> metatags or negated fav:<name> metatags.
For ordfav:<name> searches, this may be faster. ordfav: searches had a
tag match clause (`tag_index @@ 'fav:123'`) in addition to a join on the
favs table. This was redundant, and in some cases it inhibited the query
planner from choosing a more optimal plan.
Partially addresses #4652 by eliminating another place where we depended
on the fav_string.
On the posts show page, in the favorites list, show favorites according
to the order they were added to the favorites table, rather than the
order they were added to the posts's fav_string.
On most posts these should be the same, but on old posts they may be
slightly different. The IDs of the first few hundred thousand favorites
don't appear to be in chronological order. Probably the original
favorite IDs were lost and recreated by a database move at some point in
Danbooru's history. The fav_string is also inconsistent with the
favorites table in some places (one contains favorites that aren't
contained by the other), which also throws off the order.
Partially addresses #4562 by eliminating one place where we depended on
the fav_string.
Fix script to delete all invalid email addresses. In production there
were ~4000 users with invalid email addresses because we used to not do
any validation of emails during signup.
* Refactor various user limit methods to class methods from instance
methods so they can be used outside the context of a single user.
* Remove the Danbooru.config.base_tag_query_limit option.
Allow promo codes to be used during checkout if a secret promo=true url
param is passed. Allows promo codes to be offered without having the
promo code option always appear even when there aren't any active promos.
Add the following bank redirect payment methods:
* https://stripe.com/docs/payments/bancontact
* https://stripe.com/docs/payments/eps
* https://stripe.com/docs/payments/giropay
* https://stripe.com/docs/payments/ideal
* https://stripe.com/docs/payments/p24
These methods are used in Austria, Belgium, Germany, the Netherlands,
and Poland.
These methods require payments to be denominated in EUR, which means we
have to set prices in both USD and EUR, and we have to automatically
detect which currency to use based on the user's country. We also have
to automatically detect which payment methods to offer based on the
user's country. We do this by using Cloudflare's CF-IPCountry header to
geolocate the user's country.
This also switches to using prices and products defined in Stripe
instead of generated on-the-fly when creating the checkout.
Add links to the Stripe payment page and the Stripe receipt page on
completed user upgrades.
The Stripe payment link is a link to the payment details on the Stripe
dashboard and is only visible to the owner.
Redirect the old user upgrade page, /user_upgrade/new, to the new
user upgrade page, /user_upgrades/new page. Some old forum posts still
link to the old page.
Log the Referer header, as well as the Sec-Fetch-* headers. These are
only sent by recent versions of Chrome; see https://www.w3.org/TR/fetch-metadata.
Require new accounts to verify their email address if any of the
following conditions are true:
* Their IP is a proxy.
* Their IP is under a partial IP ban.
* They're creating a new account while logged in to another account.
* Somebody recently created an account from the same IP in the last week.
Changes from before:
* Allow logged in users to view the signup page and create new accounts.
Creating a new account while logged in to your old account is now
allowed, but it requires email verification. This is a honeypot.
* Creating multiple accounts from the same IP is now allowed, but they
require email verification. Previously the same IP check was only for
the last day (now it's the last week), and only for an exact IP match
(now it's a subnet match, /24 for IPv4 or /64 for IPv6).
* New account verification is disabled for private IPs (e.g. 127.0.0.1,
192.168.0.1), to make development or running personal boorus easier
(fixes#4618).
Tag.category_for looked up a tag's category in the Redis cache. This was
only used in a few places (in related tags, and on the popular/missed
search pages). Get rid of this method so we can work towards getting rid
of caching tag categories in Redis.
We used to not validate user email addresses, which means we have a lot
of users with invalid emails. This script deletes all emails that are
missing both an `@` and a `.` This amounts to about 3000 invalid emails.
There are an additional ~1000 emails that are missing just the `@` sign.
Many of these are simple typos, for example skipping the `@` or typing a
2 instead. Some of these may be manually fixable.
This fixes an issue where upgrading to Gold could fail if you had an
invalid email address, because we prefilled the buyer's email address on
the Stripe checkout page and an invalid email would cause Stripe to
throw an error.
Mark links to the login page as rel="nofollow" so that search crawlers
don't constantly try to crawl it. Otherwise the fact the login url is
different on every page (/login?url=<current_url>) confuses crawlers.
Also strip the url param from the canonical url (<link rel="canonical">)
on the login page.
evazion
evazion