This method had grown into a monster. Its query had bugs
that I couldn't fix, and new features would be hard to add.
Also I don't understand how it all works anymore...
Replace it with common table expressions that can be queried
to generate the results we need, instead of subtracting
results using lots of "NOT IN" clauses.
Fixed are bugs with tag schemas that use combinations of
tag groups, parent tags, and one-tag-per-topic restrictions.
For example: https://meta.discourse.org/t/130991/6
If there is any other whitespace in the container, then Firefox will add a leading/trailing space when double clicking the key. This commit wraps the key in a div with no whitespace, to work around the problem.
* DEV: Provide radix 10 argument to parseInt
* DEV: Provide radix 16 argument to parseInt
* DEV: Remove unnecessary parseInt calls
* Fix year formatting
parseInt was used here to convert decimals to ints
I made a regression here 17366d3bcc (diff-ddeebb36d131f89ca91be9d04c2baefaR10)
When the tag is added, people watching specific tag are notified but also people watching specific category.
Therefore, `notify_post_users` should accept options who should be notified.
So when `category` is added to the topic, users watching topic and users watching category are notified.
When `tag` is added to the topic, users watching topic and users watching tag are notified
Finally, when a new post is created, everybody is notified, topic watchers, category watchers, tag watchers.
Previous versions of the mail-receiver used query based api credentials,
if we detect this we will show a message in the admin panel to update
the mail receiver.
* Fix user title logic when badge name customized
* Fix an issue where a user's title was not considered a badge granted title when the user used a badge for their title and the badge name was customized. this affected the effectiveness of revoke_ungranted_titles! which only operates on badge_granted_titles.
* When a user's title is set now it is considered a badge_granted_title if the badge name OR the badge custom name from TranslationOverride is the same as the title
* When a user's badge is revoked we now also revoke their title if the user's title matches the badge name OR the badge custom name from TranslationOverride
* Add a user history log when the title is revoked to remove confusion about why titles are revoked
* Add granted_title_badge_id to user_profile, now when we set badge_granted_title on a user profile when updating a user's title based on a badge, we also remember which badge matched the title
* When badge name (or custom text) changes update titles of users in a background job
* When the name of a badge changes, or in the case of system badges when their custom translation text changes, then we need to update the title of all corresponding users who have a badge_granted_title and matching granted_title_badge_id. In the case of system badges we need to first get the proper badge ID based on the translation key e.g. badges.regular.name
* Add migration to backfill all granted_title_badge_ids for both normal badge name titles and titles using custom badge text.
* If a staff user created only a security key as their single 2FA option. they continued to be prompted to create a 2FA option because we only considered this condition satisfied if a TOTP was added.
* The condition is now satisfied if TOTP OR security keys are enabled.
Instead of enabling `suppress_from_latest` setting on many categories now we can enable `mute_all_categories_by_default` site setting. Then users should opt-in to categories for them to appear in the latest and categories pages.
Previously our index was non unique, causing situations where a user could
have multiple drafts stored in the table for the same exact entity.
This does not properly reflect reality and needed to change as in certain
cases duplicate drafts could be created causing internal data inconsistency
This change adds a message to the admin panel if it detects an api
requests that doesn't use the new header based authentication method.
The message is to warn people to switch to header based auth and links
to the api documention topic on meta for more info.
Issue was mentioned in this [meta topic](https://meta.discourse.org/t/send-a-notification-to-watching-users-when-adding-tag/125314)
It is working well when category is changed because NotifyCategoryChange job already got that code:
```
if post&.topic&.visible?
post_alerter = PostAlerter.new
post_alerter.notify_post_users(post, User.where(id: args[:notified_user_ids]))
post_alerter.notify_first_post_watchers(post, post_alerter.category_watchers(post.topic))
end
```
For NotifyTagChange job notify post users were missing so it worked only when your notification was set to `watching first post`
- Allow revoking keys without deleting them
- Auto-revoke keys after a period of no use (default 6 months)
- Allow multiple keys per user
- Allow attaching a description to each key, for easier auditing
- Log changes to keys in the staff action log
- Move all key management to one place, and improve the UI
These are the changes to tags routing that correspond to the category
routing changes of d84c34ad. The new scheme is:
/tags/c/*slug_path/:id/none/:tag_id/ENDPOINT
/tags/c/*slug_path/:id/none/:tag_id
/tags/c/*slug_path/:id/:tag_id/ENDPOINT
/tags/c/*slug_path/:id/:tag_id
Previously theme translations were loaded along with other plugin API scripts. These run after pre-initializers and initializers when the app boots. This commit moves theme translation loading into pre-initializers, so their behaviour matches core translations more closely.
Previously we could not place extra nav items in a particular order.
This change introduces a new `before` attribute to addNavigationBarItem
with this attribute we can place a nav item before "top" or any other place.
Additionally this allows navigation items to force active state. In some
cases we may want an items that is simply a "filter" on latest using `?query_params`
when we do that we can "force" the filter active.
A concrete example is the assign plugin that adds a nav item for
Unassigned => `/c/cat-slug/l/latest?assigned=nobody&status=open`
If we did not force then latest would be selected even though the filter
is clicked.
This ensures we can very quickly figure out which topics are banners if
a banner is set.
Previously you would have to scan an entire table to find banners
In a category's settings, the Tags tab has two new fields to
specify the number of tags that must be added to a topic
from a tag group. When creating a new topic, an error will be
shown to the user if the requirement isn't met.
The routes for categories are changing. The scheme that I intend to move
us to is:
/c/*slug_path/(:id)/ENDPOINT
/c/*slug_path/(:id)
This commit adds support for the new scheme to the server side without
dropping support for existing URLs. It is necessary to support existing
URLs for two reasons:
* This commit does not change any client side routing code,
* Posts that contain category hashtags that refer to a root category
are baked into URLs that do not fit this new scheme, (/c/[id]-[slug])
* FIX: move attachment_css_class constant out of upload-short-url for discourse-markdown-it
* Use setTimeout instead of ember later
* WIP. Not sure if this worked.
* oneboxer cache in separate file
* Reset onebox cache still
* set functions for oneboxers cache
This is a major change to draft internals. Previously there were quite a
few cases where the draft system would say "draft saved", when in fact
we just skipped saving.
This commit ensures the draft system deals with draft ownership handover in
a predictable way.
For example:
- Window 1 editing draft
- Window 2 editing same draft at the same time
Previously we would allow window 1 and 2 to just fight on the same draft
each window overwriting the same draft over an over.
This commit introduces an ownership concept where either window 1 or 2 win
and user is prompted on the loser window to reload screen to correct the issue
This also corrects edge cases where a user could have multiple browser windows
open and posts in 1 window, later to post in the second window. Previously
drafts would break in the second window, this corrects it.
* DEV: Add the actual "tag_groups/new" route
Allows refreshing the "new" page without an error.
* DEV: Prevent attempts to create group tags if tagging is disabled
* DEV: Refactor the tag-groups controller
Gets rid of `selectedItem`, `selected`, and `selectTagGroup` action.
* DEV: Rename tag-groups-show to tag-groups-edit
* DEV: Refactor tag-groups form
* Extracted the tag-groups-form that's used by tag-groups-new and tag-groups-edit
* The model is now a buffered property
* Serialization relies more heavily on RestAdapter now
* Data is sent as JSON
* Payload is now namespaced ("tag_group")
* Update app/assets/javascripts/discourse/controllers/tag-groups-new.js.es6
Co-Authored-By: Joffrey JAFFEUX <j.jaffeux@gmail.com>
* Update app/assets/javascripts/discourse/components/tag-groups-form.js.es6
Co-Authored-By: Joffrey JAFFEUX <j.jaffeux@gmail.com>
* Update app/assets/javascripts/discourse/controllers/tag-groups-edit.js.es6
Co-Authored-By: Joffrey JAFFEUX <j.jaffeux@gmail.com>
* Require q param in /tags/filter/search route.
* If not provided this route was causing a 500 error when
DiscourseTagging.clean_tag was called, because .downcase
was being called on the param (which was nil).
* Now return a 400 error instead.
Adds the settings:
raw_email_max_length, raw_rejected_email_max_length, delete_rejected_email_after_days.
These settings control retention of the "raw" emails logs.
raw_email_max_length ensures that if we get incoming email that is huge we will truncate it removing uploads from the raw log.
raw_rejected_email_max_length introduces an even more aggressive truncation for rejected incoming mail.
delete_rejected_email_after_days controls how many days we will keep rejected emails for (default 90)
* DEV: Refactor setting component save callback
* refactor site-setting component around new callback
* add callback to theme-translation component
* remove the save callback altogether
Anonymous users are only possible if allow_anonymous_posting is true,
which means that 'user.is_anonymous' check implies that
allow_anonymous_posting is true.
The server already ensures it advances draft keys when a post is created
this means this code that used to delete drafts is simply introducing
composer delays with no benefit.
Defer placing scale buttons causes "scale image" buttons to be missing if
you manage to scroll mouse to the area where the image is prior to image
rendering.
This fix ensures scale buttons are always rendered leaving all the hiding
and styling entirely to CSS.
This also corrects a bug where scaling toolbar was missing on mobile
* FEATURE: Site setting/ui to allow users to set their primary group
* prettier and remove logic from account template
* added 1 to 43 to make web_hook_user_serializer_spec pass
Previously every hour we would run a full scan of the entire DB searching
for expired uploads that need to be moved to the tombstone folder.
This commit amends it so we only run the job 2 times per clean_orpha_uploads_grace_period_hours
There is a upper bound of 7 days so even if the grace period is set really
high it will still run at least once a week.
By default we have a 48 grace period so this amends it to run this cleanup
daily instead of hourly. This eliminates 23 times we run this ultra expensive
query.
The query to count how many new users there are since a given date
is expensive. It's the least personalized stat and the one we fallback
to last when no better number can be found for the target user.
Give up accuracy so we can aggressively cache the user counts
that appear in this email.
Certain DBs have duplicates already, if we want to ensure uniqueness here
we need to decide first how to clean up existing data and confirm all the
plugins expect this.
* FEATURE: Add remembering topic list for group pms
* added findOrResetCachedBy helper in topic-list
* Created cached-topic-list.js
* Update app/assets/javascripts/discourse/routes/build-private-messages-route.js.es6
Co-Authored-By: Robin Ward <robin.ward@gmail.com>
This simplifies the code for refreshing notification counts. It now
unconditionally looks up the user object which protects against stale objects
in memory.
This also removes a pile of conditional logic we no longer need.
This fix ensures that searches that contain a null byte return a 400
error instead of causing a 500 error.
For some reason from rspec we will reach the raise statement inside
of the `rescue_from ArgumentError` block, but outside of rspec it will
not execute the raise statement and so a 500 is thrown instead of
reaching the `rescue_from Discourse::InvalidParameters` block inside of
the application controller.
This fix raises Discourse::InvalidParameters directly from the search
controller instead of relying on `PG::Connection.escape_string` to
raise the `ArgumentError`.
The payload when receiving a notification webhook is pointless without
knowing which user the notification is for. This fix adds the user_id to
the notification serializer so that when you receive a notification
webhook you can properly identify which user the notification is for.
See
https://meta.discourse.org/t/getting-the-target-user-for-notification-webhook-events/129052?u=blake
for more details.
When autocompleting mentions in secure categories, we immediately populate the list with users which have permission to view the category. This logic is applied to unsecured categories as well, but the server returns an empty list of users. This commit teaches the autocomplete to understand empty lists of users without terminating the autocomplete dropdown.
Doing .pluck(:column).first is a very common pattern in Discourse and in
most cases, a limit cause isn't being added. Instead of adding a limit
clause to all these callsites, this commit adds two new methods to
ActiveRecord::Relation:
pluck_first, equivalent to limit(1).pluck(*columns).first
and pluck_first! which, like other finder methods, raises an exception
when no record is found
- destroyDraft which is called when we cancel a draft is now async,
removing race conditions when you click "reply" to a post and are
already editing. We used to trigger double dialogs for cancelling
drafts which was confusing.
- Remove reply as new topic / reply as pm keys, they are no longer
used and only caused confustion. For example we used to pop up a
warning when you are composing a reply and flick to reply as
new topic
- Remove createTopic key, this was a bug that proliferated. Whenever
creating a topic via the C shortcut or clicking on new topic on full
screen search the correct new topic draft key will be used
consistently
- When abandoning an edit we now say "Are you sure you want to discard
your changes" (instead of abandon your post which is confusing)
- Increase size of the reviewable's conversation excerpt to prevent truncation of the new copy
- Remove the `domain` parameter from the `flag_linked_posts_as_spam` method in the user model since it is no longer needed
- Remove the `domain` interpolation variable from all translation files
- Add "All posts from this user that include links should be reviewed." to server.en.yml for added clarity on why the posts entered the queue
* DEV: Remove badge-select-controller
1. The `selectableUserBadges` computed property of `badge-title` was being incorrectly overwritten, which triggered computed-property.override deprecation.
2. The `badge-select-controller` mixin contained properties that were used either by `badge-title` component or `badges/show` controller, but none were being used by both. This change moves properties where they belong, and removes the mixin.
* Update app/assets/javascripts/discourse/controllers/badges/show.js.es6
Co-Authored-By: Robin Ward <robin.ward@gmail.com>
* Improve code brevity
`suspend` isn't a User attribute, but was being assigned to the frontend User model as if it was. The model has a computed property that depends on `suspended_till`, so instead of overriding this property, it's better to return relevant attributes.
Fixes a computed-property.override deprecation (https://emberjs.com/deprecations/v3.x#toc_computed-property-override)
The 'automatically set primary group' checkbox looked like it was associated with the email membership. In fact, it applies to all members who join the group. This commit moves it next to the 'automatic trust level' setting, and puts them both under an 'Effects' heading
Under exceptional situations the automatic draft feature can fail.
This new **hidden, default off** site setting
`backup_drafts_to_pm_length` will automatically backup any draft that is
saved by the system to a dedicated PM (originating from self)
The body of that PM will contain the text of the reply.
We can enable this feature strategically on sites exhibiting issues to
diagnose issues with the draft system and offer a recourse to users who
appear to lose drafts. We automatically checkpoint these drafts every 5
minutes forcing a new revision each 5 minutes so you can revert to old
content.
Longer term we are considering automatically enabling this kind of feature
for extremely long drafts where the risk is really high one could lose
days of writing.
* FIX: Do not encode the URL twice
Now that we encode slugs in the server we don't need this anymore.
Reverts fe5na33
* FIX: More places do deal with encoded slugs
* the param is a string now, not a hash
* FIX: Handle the nil slug on /categories
* DEV: Add seeded? method to identity default categories
* DEV: Use SiteSetting to keep track of seeded categories
Slugs can be the empty string, but the added index didn't account for
that. This commit changes the migration, stopping it from being unique
so that it can be applied everywhere and adds another migration that
recreates the index properly.
This feature amends it so instead of using one challenge and honeypot
statically per site we have a rotating honeypot and challenge value which
changes every hour.
This means you must grab a fresh copy of honeypot and challenge value once
an hour or account registration will be rejected.
We also now cycle the value of the challenge when after successful account
registration forcing an extra call to hp.json between account registrations
Client has been made aware of these changes.
Additionally this contains a JavaScript workaround for:
https://bugs.chromium.org/p/chromium/issues/detail?id=987293
This is client side code that is specific to Chrome user agent and swaps
a PASSWORD type honeypot with a TEXT type honeypot.
When a category has a subcategory, we ensure that no one who can see the
subcategory cannot see the parent. However, we don't take into account
the fact that, when no CategoryGroups exist, the default is that
everyone has full permissions.
Moving posts also moves the read state (`topic_users` table) to the destination topic. This changes that behavior so that only users who posted in the destination topic will have the original notification level (probably "watching") of the original topic. The notification level for all other users will be set to "regular".
When an admin changes the site setting slug_generation_method to
encoded, we weren't really encoding the slug, but just allowing non-ascii
characters in the slug (unicode).
That brings problems when a user posts a link to topic without the slug, as
our topic controller tries to redirect the user to the correct URL that contains
the slug with unicode characters. Having unicode in the Location header in a
response is a RFC violation and some browsers end up in a redirection loop.
Bug report: https://meta.discourse.org/t/-/125371?u=falco
This commit also checks if a site uses encoded slugs and clear all saved slugs
in the db so they can be regenerated using an onceoff job.
* DEV: allow serializing data for drafts
* Various fixes
* added an alias method for 'serializeToDraft' to plugin-api
* fixed linting issues
* changed single quotes to double quotes to fix linting issue
* fixed linting issues
* fixed composer model file via prettier
* fixed composer controller file via prettier
* fixed plugin-api file via prettier
Our instance used for template rendering needs a lock to ensure there is
no race condition where rendering happens on 2 threads at the same time.
This can lead to local poisoning which can cause unexpected results in
emails
A modal's primary action (blue button in the default theme) can now be invoked
by hitting Enter on the keyboard. This applies to all modals that aren't strict
forms as long as the focus is not on a textarea element.
Dropping the temp table in an `ensure` block hides the actual exception. Creating the table with `ON COMMIT DROP` makes the temp table disappear automatically at the end of the transaction. We only need the explicit `DROP` in tests, because tests already run inside a transaction, so the temp table won't be dropped after each test which leads to spec failures.
Bump onebox version, and add new styling
Commit, PR and Issue oneboxes are updated with a new design. Timestamps are now localized using local-dates (if installed).
We were mixing in 20 or so methods into a controller just to use a single
one.
The helper itself is not the actual implementation anyway... MobileDetection
is responsible here.
We expect mini profiler only to show up on accounts that are flagged as
developer accounts.
Unfortunately there was a bypass on any controllers that mix in ApplicationHelper
Post timings are created by `topic_id` and `post_number` and it's possible that the destination topic already contains post timings for non-existent posts. For example, this can happen if the destination topic was previously split and Discourse recorded post timings for moved posts in the destination topic.
This commit ensures that all timings which reference non-existent posts are deleted from the destination topic before the posts are moved.
This addresses the following issues:
- on iPad, with keyboard attached, the composer is no longer forced to full screen
- on iPad, with keyboard attached, the topic no longer scrolls when starting a
reply and then cancelling it
- switching between inputs and buttons (formatting, emojis, categories/tags, etc.) no longer
causes layout to bounce around
If the setting is turned on, then the user will receive information
about the subject: if it was deleted or requires some special access to
a group (only if the group is public). Otherwise, the user will receive
a generic #404 error message. For now, this change affects only the
topics and categories controller.
This commit also tries to refactor some of the code related to error
handling. To make error pages more consistent (design-wise), the actual
error page will be rendered server-side.
Using popups is becoming increasingly rare. Full page redirects are already used on mobile, and for some providers. This commit removes all logic related to popup authentication, leaving only the full page redirect method.
For more info, see https://meta.discourse.org/t/do-we-need-popups-for-login/127988
Removes setting for iOS devices that support Visual Viewport API.
On devices where it was previously enabled, it was causing some scrolling drift when invoking the composer.
This is useful by analytics libraries or other code that wants to track
when replies are begun. A new event: `page:compose-reply` is fired with
the topic.
If we are interested in page events (say analytics), they are reported
when the route changes, which does not wait for any promises in
`setupController` to finish.
A plugin might want to know when a topic has fully loaded, so this event
is triggered when that happens.
Bump onebox version to include new github rendering, and add relevant CSS
Avatars are reduced in size significantly, and icons are added to easily differentiate PRs and commits. The 'Issue:' prefix is removed from issue oneboxes, to make them consistent with commits and PRs.
Previous to this fix we were leaking methods on the internal action view
template class per render.
This caused email generation to be very low and a steady memory leak in the
application in sidekiq when sending out emails
The behavior change is new to Rails 6 so this fix does not need to be
backported into stable.
* FEATURE: Added input for name when creating a new authenticator in user preferences
* FEATURE: Added placeholders to authenticator inputs
* Ran prettier on second-factor-add-totp.js.es6
AppEvents was always a service object in disguise, so we should move it
to the correct place in the application. Doing this allows other service
objects to inject it easily without container access.
In the future we should also deprecate `this.appEvents` without an
explicit injection too.
Added on Aug 21, 2015 in bef80633b1
The only usage removed on Aug 26, 2015 in 4ba89eec27
As far I can tell it isn't used by core or any official plugins.
* FEATURE: Adds an extra protection layer when decompressing files.
* Rename exporter/importer to zip importer. Update old locale
* Added a new composite class to decompress a file with multiple strategies
* Set max file size inside a site setting
* Ensure that file is deleted after compression
* Sanitize path and files before compressing/decompressing
This fix is needed due to what feels like an iOS Safari bug. The CSS rule `margin-bottom: env(safe-area-inset-bottom);` should not apply to the topic progress element when the composer is visible, because the element is not near the bottom of the viewport.
* Fix broken security key 2FA on mobile login.hbs
* Show nicer error message when a security key already exists
* [COPY] Disable -> Delete for security key editing
* Standardize UI elements in 2FA prefs password confirmation
* Minor fixes to label location for resetPasswordProgress
Partially reverts 94ab48c by using Safari hacks on iPad again.
This brings parity in the composer UI between iPhones and iPads
Hides grippie and fullscreen toggle button when the keyboard is visible on iPads
Clicking fast on the "top", "unread", or "latest" button when browsing a parent category page with subcategories and the setting `Show subcategory list above topics in this category` enabled would cause an exception:
```
Uncaught Error: Nothing handled the action 'triggerRefresh'. If you did handle the action, this error can be caused by returning true from an action handler in a controller, causing the action to bubble.
```
Zeitwerk simplifies working with dependencies in dev and makes it easier reloading class chains.
We no longer need to use Rails "require_dependency" anywhere and instead can just use standard
Ruby patterns to require files.
This is a far reaching change and we expect some followups here.
Adds 2 factor authentication method via second factor security keys over [web authn](https://developer.mozilla.org/en-US/docs/Web/API/Web_Authentication_API).
Allows a user to authenticate a second factor on login, login-via-email, admin-login, and change password routes. Adds registration area within existing user second factor preferences to register multiple security keys. Supports both external (yubikey) and built-in (macOS/android fingerprint readers).
Currently, the topic is only validated for censored words and should be validated for blocked words as well.
Blocked word validation is now used by both Post and Topic. To avoid code duplication, I extracted blocked words validation code into separate Validator, and use it in both places.
The only downside is that even if the topic contains blocked words validation message is saying "Your post contains a word that's not allowed: tomato" but I think this is descriptive enough.
Some site settings (e.g. `unicode_username_character_whitelist`) depend on the default locale, so we need to reset the watched settings when the locale changes.
Keeps element 1em away from the right edge of screen
Takes DiscourseHub app nav position into account on iPad
Uses outerHeight to calculate element height including padding/borders
Per new lifecycle https://developers.google.com/web/updates/2018/07/page-lifecycle-api
On Android and latest Chrome when an app transitions from "frozen" to
active the new "resume" event fires with no accompanying "visibilitychange"
event.
This means that often background tabs may be stuck thinking that discourse
has no focus when, indeed, it has.
This leads to cases where no posts are marked read anymore.
This applies to iPhones running iOS 13+.
Previous technique remains in place for iOS 12 and below.
Note that this does not apply to iPads on iOS 13 due to Apple no longer
identifying iPads in the user agent string.
The current manifest validation in Chrome requires at least one
non-maskable icon to make it installable.
This commit adds the maskable entry as another entry, following up
on changes added in 3e590b8
Because:
- Chrome 78 and Firefox Fenix have support for it
- The icons will look better by using all the availiable space
- Admins can control the icon and add appropriate padding since we have a
dedicated asset for the manifest logo. Read more about it on
https://css-tricks.com/maskable-icons-android-adaptive-icons-for-your-pwa/
This commit:
- Adds a new key under the icons array in the webmanifest, named purpose
with maskable value.
Previously, calculating thresholds for reviewables was done based on the
50th and 85th percentile across all reviewables. However, many forum
owners provided feedback that these thresholds were too easy to hit, in
particular when it came to auto hiding content.
The calculation has been adjusted to base the priorities on reviewables
that have a minimum of 2 scores (flags). This should push the amount of
flags required to hide something higher then before.
On forums with very few flags you don't want to calculate averages
because they won't be very useful. Stick with the defaults until we hit
15 reviewables at least.
Forums without previously calculated scores would return the same values
for low/medium/high sensitivity. Now those are scaled based on the
default value.
The default value has also been changed from 10.0 to 12.5 based on
observing data from live discourse forums.
We were counting all the oneboxes in the DOM instead of just the ones in the preview.
Also refactored the logic to count up to 'max_oneboxes_per_post` instead of down to 0.
That also ensured we don't load 11 oneboxes when the setting is limiting to 10.
The dollar sign (`$`) is a special replace pattern, and `$&` inserts the
matched string. Thus dollars signs need to be escaped with the special
pattern `$$`, which inserts a single `$`.
In development, we track the last requested theme id, and use that to refresh the correct stylesheet targets. The after_action hook runs on every request, but the preview_theme_id parameter is only sent on the initial HTML request. This commit ensures we only fetch the development theme_id on HTML requests
To demonstrate the issue:
- Visit https://meta.discourse.org/#somethingHere while logged in
- Click "log out"
- You will be logged out, but the page will not be reloaded
Setting `window.location.pathname = "/"` will not reload the page if there is a hash present. Using `window.location = "/"` gives us the desired behavior.
In IE11, the browser returns the cached HTML response, rather than the JSON formatted response. A better solution may be to add a `Vary: Accept` header to all of our HTML responses, but this commit should solve the immediate issue.
`fancy_title` is already escaped by Rails. Escaping it again would print
the HTML entity as-is, e.g. `"` instead of `"`.
This fixes the issue by introducing a new `escapedContent` attribute on
the `QuickAccessItem` widget.
* FIX: Cast all numerical values in reports
The backend can return some numerical values in report as strings. That results in unexpected order of values when sorting report tables.
* Create `toNumber()` helper
The `typeof` and `parseFloat` seem to be the fastest path: https://jsperf.com/number-vs-typeof-vs-parsefloat#results
* Adjustments to pass specs on Rails 6.0.0
* Use classic autoloader instead of Zeitwerk
* Update Rails 6.0.0 deprecated methods
* Rails 6.0.0 not allowing column with integer name
* Drop freedom_patches/rails6.rb
* Default value for trigger_transactional_callbacks? is true
* Bump rspec-rails version to 4.0.0.beta2
* FIX: inline_uploads and subfolder
* if subfolder, also look for images with a path containing
cdn_url + relative_url_root
* FIX: migrate_to_s3 task and subfolder
New site setting: `embed_any_origin` that will send postMessages to
wildcard origins `*` instead of the referer.
Most of the time you won't want to do this, so the setting is default to
`false`. However, there are certain situations where you want to allow
embedding to send post messages when there is no HTTP REFERER.
For example, if you created a native mobile app and you wanted to embed a list
of Discourse topics as HTML. In the code your HTML would be a
static file/string, which would not be able to send a referer. In this
case, the site setting will allow the embed to work.
From a security standpoint we currently only use `postMessage` to send
data about the size of the HTML document and scroll position, so it
should be enable if required with minimal security ramifications.
* Extract QuickAccessPanel from UserNotifications.
* FEATURE: Quick access panels in user menu.
This feature adds quick access panels for bookmarks and personal
messages. It allows uses to browse recent items directly in the user
menu, without being redirected to the full pages.
* REFACTOR: Use QuickAccessItem for messages.
Reusing `DefaultNotificationItem` feels nice but it actually requires a
lot of extra work that is not needed for a quick access item.
Also, `DefaultNotificationItem` shows an incorrect tooptip ("unread
private message"), and it is not trivial to remove / override that.
* Use a plain JS object instead.
An Ember object was required when `DefaultNotificationItem` was used.
* Prefix instead suffix `_` for private helpers.
* Set to null instead of deleting object keys.
JavaScript engines can optimize object property access based on the
object’s shape. https://mathiasbynens.be/notes/shapes-ics
* Change trivial try/catch to one-liners.
* Return the promise in case needs to be waited on.
* Refactor showAll to a link with href
* Store `emptyStatePlaceholderItemText` in state.
* Store items in Session singleton instead.
We can drop `staleItems` (and `findStaleItems`) altogether. Because
`(old) items === staleItems` when switching back to a quick access
panel.
* Add `limit` parameter to the `user_actions` API.
* Explicitly import Session instead.
This reverts commit 310a8ac242.
It seems this breaks google authentication. My suspicion is opening
the URL twice invalidates the CSRF after the first access.
* FEATURE: Add tl2 threshold for editing new posts
* Adds a new setting and for tl2 editing posts (30 days same as old value)
* Sets the tl0/tl1 editing period as 1 day
* FIX: Spec uses wrong setting
* Fix site setting on guardian spec
* FIX: post editing period specs
* Avoid shared examples
* Use update_columns to avoid callbacks on user during tests
