* Introduce domains method to Account relation
Account had followers_domains method, which was excessively specific.
Let relation of Account have domains method instead.
* Move follow_mapping in Account to AccountInteractions
* Introduce shared examples for AccountAvatar inclusion
* Cover Account more
* Fix regression from #3842
Simplify the query by omitting all direct statuses. Private statuses
are allowed because they are from accounts we are following (so
by definition)
Resolves#3887 (alternative)
* Adjust test
(This patch has been merged as bugfix and reverted, but still valuable as
improvement)
Previously, we've attached IntersectionObserver twice for boosted statuses:
wrapper Status and wrapped Status. but wrapped Status don't need to manage
intersection and visibility by itself, because it's a part of wrapper Status.
* Revert "Bump version to 1.4.4"
This reverts commit 1585b0c6cc.
* Revert "Fix conversations (fixes#3869) (#3870)"
This reverts commit 15b43f555d.
* Revert "Fix streaming server. Redis connection subscribe for each channel. (#3828)"
This reverts commit d8ec832806.
* Revert "Filter direct statuses in Status.as_home_timeline (#3842)"
This reverts commit bab5a18232.
* Revert "Fix RemoteFollow behavior (#3868)"
This reverts commit a20cf3b64e.
* Revert "Update fabricator for MediaAttachment to attach a file according to type (#3862)"
This reverts commit 356df7ae6b.
* Revert "Upgrade React Router (#3677)"
This reverts commit 8f03fdce7f.
* Revert "Do not call setState from unmounted component (#3853)"
This reverts commit 1fc6cb4997.
* Revert "Replace TextIconButton for SensitiveButton to IconButton (#3759)"
This reverts commit eb832e88f4.
* Revert "Fix RTL detection on Ruby side (#3867)"
This reverts commit b16b69350e.
* Revert "i18n: Fixed typo in Polish translation (#3864)"
This reverts commit da6fa029f6.
* Revert "Don't attach IntersectionObserver for wrapped statuses (#3863)"
This reverts commit 94ad0706f5.
The classes using Status.as_home_timeline, namely Feed and
PrecomputeFeedService are expected to filter direct statuses as
FanOutWriteService does, but their filtering were incomplete or missing.
This commit solves the problem by filtering direct statuses in
as_home_timeline as the other similar methods such as as_public_timeline
does.
This fixes a bug that sometimes boosted statuses being hidden on scrolling.
Previously, we've attached IntersectionObserver twice for boosted statuses:
wrapper Status and wrapped Status. This will call intersection handler twice,
so this may results race condition...probably.
* Whitelist allowed classes for federated statuses
Allowed classes are currently:
- Any microformats class (h/p/u/dt/e-*)
- the classes mention, hashtag, ellipses and invisible.
this last one is somewhat suspect, but Mastodon currently uses it to render hidden link text.
resolved#3790
* Fix code style
mergeDeep also merges columns, but it should be replaced simply.
So in the new function, first apply mergeDeep except columns, and set default columns if columns unset.
* Make Pubsubhubbub::DistributionWorker handle both single stream entry
arguments, as well as arrays of stream entries
* Add BatchedRemoveStatusService, make SuspendAccountService use it
* Improve method names
* Add test
* Add more tests
* Use PuSH payloads of 100 to have a clear mapping of
1000 input statuses -> 10 PuSH payloads
It was nice while it lasted
* Add form for account deletion
* If avatar or header are gone from source, remove them
* Add option to have SuspendAccountService remove user record, add tests
* Exclude suspended accounts from search
* Fix#2619 - When redis feed is empty, fall back to database
* Use redis value to return feed from database only while RegenerationWorker
hasn't finished running
* Fix specs
* Replace usage of reject!
TagManager.local_url? was sometimes called with an URI with a nil host,
leading to a crash in TagManager.local_url?. This fixes moves the
already-existing uri.host.blank? check in front to avoid this case.
* Move ancestors/descendants out of timelines reducer
* Refactor timelines reducer
All types of timelines now have a flat structure and use the same
reducer functions and actions
* Reintroduce some missing behaviours
* Fix wrong import in reports
* Fix includes typo
* Fix issue related to "next" pagination in timelines and notifications
* Fix bug with timeline's initial state, expandNotifications
```
DEPRECATION WARNING: The behavior of `attribute_changed?` inside of after callbacks will be changing in the next version of Rails. The new return value will reflect the behavior of calling the method after `save` returned (e.g. the opposite of what it returns now). To maintain the current behavior, use `saved_change_to_attribute?` instead. (called from block in <class:Account> at /Users/rene/Workspace/personal/ruby/mastodon/app/models/account.rb:60)
DEPRECATION WARNING: The behavior of `attribute_changed?` inside of after callbacks will be changing in the next version of Rails. The new return value will reflect the behavior of calling the method after `save` returned (e.g. the opposite of what it returns now). To maintain the current behavior, use `saved_change_to_attribute?` instead. (called from block in <class:Account> at /Users/rene/Workspace/personal/ruby/mastodon/app/models/account.rb:60)
DEPRECATION WARNING: The behavior of `attribute_changed?` inside of after callbacks will be changing in the next version of Rails. The new return value will reflect the behavior of calling the method after `save` returned (e.g. the opposite of what it returns now). To maintain the current behavior, use `saved_change_to_attribute?` instead. (called from block in <class:Account> at /Users/rene/Workspace/personal/ruby/mastodon/app/models/account.rb:60)
DEPRECATION WARNING: The behavior of `attribute_changed?` inside of after callbacks will be changing in the next version of Rails. The new return value will reflect the behavior of calling the method after `save` returned (e.g. the opposite of what it returns now). To maintain the current behavior, use `saved_change_to_attribute?` instead. (called from block in <class:Account> at /Users/rene/Workspace/personal/ruby/mastodon/app/models/account.rb:61)
DEPRECATION WARNING: The behavior of `attribute_changed?` inside of after callbacks will be changing in the next version of Rails. The new return value will reflect the behavior of calling the method after `save` returned (e.g. the opposite of what it returns now). To maintain the current behavior, use `saved_change_to_attribute?` instead. (called from block in <class:Account> at /Users/rene/Workspace/personal/ruby/mastodon/app/models/account.rb:62)
DEPRECATION WARNING: The behavior of `attribute_changed?` inside of after callbacks will be changing in the next version of Rails. The new return value will reflect the behavior of calling the method after `save` returned (e.g. the opposite of what it returns now). To maintain the current behavior, use `saved_change_to_attribute?` instead. (called from block in <class:Account> at /Users/rene/Workspace/personal/ruby/mastodon/app/models/account.rb:63)
```
Here's PR describing changes to Dirty API https://github.com/rails/rails/pull/25337
When case insensitivity is enabled via devise's `config.case_insensitive_keys` then `.find_for_authentication` method needs to be used instead of `.find_by` because second mentioned returns `nil` when valid email with different cases is passed.
More info https://github.com/plataformatec/devise/wiki/How-To:-Use-case-insensitive-emails
- Use plaintext
- Strip out URLs
- Strip out mentions
- Strip out hashtags
- Strip out whitespace from "overall" count
- Consistent between JS and Ruby
* Improve default language decision
This change allows to takes account of accepted language determined by
the user agent even if the custom default locale of the instance is
configured.
* Cover Localized more
* Fix code style
Each of mute, favourite, reblog has been updated to:
- Have a separate controller with just a create and destroy action
- Preserve historical route names to not break the API
- Mild refactoring to break up long methods
* Add specs for api statuses routes
* Update favourited_by and reblogged_by api routes
* Move methods into new controllers
* Use load_accounts methods to simplify index actions
* Clean up load_accounts methods
* Clean up link header generation
* Check for link headers in specs
* Remove unused actions from api/v1/statuses controller
* Remove specs for moved actions
* Move ApiController to Api/BaseController
* API controllers inherit from Api::BaseController
* Add coverage for various error cases in api/base controller
* Coverage for rate limit headers
* Move rate limit headers methods to concern
* Move throttle check to condition on before_action
* Move match_data variable into method
* Move utc timestamp to separate method
* Move header setting into smaller methods
* specs cleanup
Steps to reproduce the original issue:
1. Have two remote accounts, A that you don't follow, and B that you follow.
2. Have A post a toot and reply to it.
3. Boost A's reply from remote account B.
This used to cause the local instance to get A's reply but fail to link it to
the original post.
* Add regex filter on the community timeline and the public timeline
* correcting
* Adjust the height of header buttons
* Remove trailing spaces
* Remove trailing spaces
* Solve some code duplication
* reset the state of the locale files in app/javascript/mastodon/locales
* adjust to upstream
* adjust to upstream
* change keys of locale settings
* Sort results by the name
* Switch search method to simple `LIKE` matching instead of tsvector/tsquery
Previously we used scores from ts_rank_cd() to sort results, but it didn't work
because the function returns same score for all results. It's not for calculate
similarity of single words. Sometimes this bug even push out exact matching tag
from results.
Additionally, PostgreSQL supports prefix searching with standard btree index.
Using it offers simpler code, but also less index size and some speed.
* Try fixing ThreadResolveWorker calls
From my understanding of ActiveRecord, a transaction is commited as soon as
the exit of the outmost ActiveRecord.transaction block. However, inner
transaction blocks will exit without the transaction being commited.
In this case, ThreadResolveWorker were fired *within* a transaction block,
so moving the call out of it should do the trick. However, this is somewhat
fragile, as this whole codepath could be called within yet another transaction.
* Set status thread within the transaction block if it is immediately available from database
* Add a StatusFilter class to identify visibility of statuses by accounts
* Extract StatusThreadingConcern from Status
* Clarify purpose of checking for nil account
* Redirect to streaming_api_base_url
When Rails receives a request to streaming API, it most likely
means that there is another host which is configured to respond
to it. This is to redirect clients to that host if
`STREAMING_API_BASE_URL` is set as another host.
* Use the new Ruby 1.9 hash syntax
* i18n Update : Add preference setting for delete toot modal
Adding a line for "Add preference setting for delete toot modal"
* i18n update for pin/unpin
Update to add two more translations
* i18n update to have the dates in plain occitan
* Removed the blank line
* %{selft} back in the translation
* Do not fall back to StreamEntry if object_type is unavailable in TagManager
Since 6d6a429af8, when Status, the only model
with stream_entry, and StreamEntry got its own logic in uri_for and
url_for, the purpose of the fallbacks to activity_type of StreamEntry
became unclear.
This commit removes the fallbacks. When adding another model with
stream_entry in future, consider to update uri_for and url_for.
* Cover TagManager more
* Do not default the format in ProviderDiscovery
The format should be determined when discovering, as it is in the current
implementation, and it is a flaw if it is not determined.
* Spec ProviderDiscovery
* Allow mounting arbitrary columns
* Refactor column headers, allow pinning/unpinning and moving columns around
* Collapse animation
* Re-introduce scroll to top
* Save column settings properly, do not display pin options in
single-column view, do not display collapse icon if there is
nothing to collapse
* Fix one instance of public timeline being closed closing the stream
Fix back buttons inconsistently sending you back to / even if history exists
* Getting started displays links to columns that are not mounted
* Add redis key "subscribed:timeline:#{account.id}" to indicate active streaming API listeners exists.
* Add endpoint for notification only stream.
* Run PushUpdateWorker only for users uses Streaming API now.
* Move close hander streamTo(Http/Ws) -> stream(Http/Ws)End (Deal with #3370)
* Add stream type for stream start log message.
* zh-cn.json: "additional info" -> "more info"
This commit changes the "extended info" (about/more) text to something that translates to "more info", as "additional info" in zh can sound like appending things to the (ugh) navbar, I guess.
Or should I just change it to "about this site" (关于本站)?
* zh-cn.json: navbar.info -> "about this site"
* zh-臺灣.json: navbar.info -> "about this site"
This implementation is a bit smaller and still has the following benefits:
* No need of app/javascript/packs/custom.js
For custom stylesheet, it typically has only
"require('../styles/custom.scss')" and is redundant.
* No need to extract vendor stylesheet to another asset
Extracting vendor stylesheet could be forgotten by developers who do not
use custom stylesheet.
* Update rails to version 5.1.1
* Run `rails app:update`
* Remove the override of polymorphic activity relationship
* Silence warning about otp_secret attribute being unknown to rails
* We will only introduce form_with where we want to use remote data
* Add failing specs for hashtag and username extraction in language detector
* Remove usernames and hashtags from text before language detection
* Handle multiple instances of special case, and reduce whitespace
* Remove trailing whitespace in i18n mailers
* Use query methods instead of #present? on AR attributes
* Delegate Status#account_domain method
* Delegate Mention #account_username and #account_acct methods
* Move specs for account finder methods to concern spec
* Move account finder methods to concern
* Improve spec wording
* Use more explicit comparison to ensure correct return value
* Add coverage for .find_local! and .find_remote!
* Add some methods to the finder
* Use arel on matching_username method
* Avoid ternary in matching domain method
* Simplify finder methods
* Use an AccountFinder class to simplify lookup
* Set delete_modal preference to true by default
* Does not show confirmation modal if delete_modal is false
* Add ja translation for preference setting page
The combination of object-fit, relative position 50% from top and translating it
back upwards 50% is what allows us to crop the video properly, so it needs to
be +50%-50%
* Fix#2922 - Load stylesheet from "custom.css" entrypoint when present
This is pretty much the same way it worked as before, albeit with
having to create app/javascript/packs/custom.js with
require('../styles/custom.scss') (or whatever you want really), which
will be a blank slate for you to import whatever you want
* Remove old assets directory
* Extract font-awesome into common.css and always load it
* Update nl strings for 1.4
* Update nl strings for 1.4
* Update nl strings for 1.4
* nl strings (+1)
More new OTP strings will be translated another time
* Fix nil input not handled well in AuthorExtractor concern
* Fix hard error in ProcessFeedService when replied-to status has been deleted
* Fix nil errors in ProcessInteractionService when favourited status
cannot be found
because it may causes flicker on the conversation when it contains blocked/muted user's status.
We use `/api/v1/statuses/{id}/context` to obtain status ids in the
conversation which filters blocked/muted user, but also uses internal
cache constructed from `in_reply_to_id` by `normalizeStatus()` in
`reducers/timelines.js` on each status loading which doesn't filter.
So statuses appears in conversation if those are cached, even those
statuses are from blocked/muted user. Then context cache will be updated
with the result of the context API and those statuses will be removed.
I have left the `normalizeStatus()` function itself which is called many
functions in the file as a placeholder for now, but maybe it should be
removed completely.
In single user mode, visitors are redirected to the single user's
profile page. So, if you are the owner without a session, you start
from that page, click the login button and authenticate yourself
expecting you'll soon get started with the home page, but in reality
you'll get redirected back to where you started from -- your own
profile page.
This fixes the behavior by redirecting you home after login if you
have started from your own profile page.
I've found this issue when I clicked replies to muted user on the timeline.
Properties I've removed in here were added with lazy loading using
IntersectionObserver (8e4d1cba), but those statuses are not need to be
tracked anyway because it will be rendered as only empty div.
This will reduce requests on who have only few statuses.
- Use next link header to detect more items from first request
- Omit next link header if result items are fewer than requested count
(It had omit it only if result was empty before)
* @object is not needed
* Remove unneeded dependencies
* Do not call private method
* Prefer #respond_to_missing? over #respond_to?
`#respond_to?` doesn't support `User.settings.method(:method_name)`
* Use find_or_initialize_by instead of
* Add load more button for large screens
* Fix `next` state value on the first loading
* Don't load if `isLoading || !hasMore`
* Start load on near the bottom
Link headers in following/followers API should include follow_id as max_id/since_id.
However, these API use current_user's account_id instead of follow_id from #3167.
This causes irrelevant result on loading more users.
- Increase coverage to exercise all parts of each action
- Move into namespace to share common code
- Misc refactor of each action for smaller methods, simpler code
* Introduce recent scope to Status and StreamEntry
Introduce recent scope to Status and StreamEntry as Account has.
* Cover AccountsController more in AccountsController
* Only load Intl data for current language
* Extract common chunk only from application.js and public.js
* Generate locale packs, avoid caching on window object
* Persian translation update
* Persian translation update: new files
* Persian translation update
* activerecord.fa.yml language code
* Persian translation update
* fix indent
* refactor(components/status_list): Avoid quering scrollTop if not necessary
* refactor(components/dropdown_menu): Do not render items if not expanded
* refactor: Cherry-pick react-motion imports
* refactor(compose/privacy_dropdown): Do not render options if not open
* refactor(components/column_collapsable): Do not render children if collapsed
* Add buttons to block and unblock domain
* Relationship API now returns "domain_blocking" status for accounts,
rename "block entire domain" to "hide entire domain", fix unblocking domain,
do not block notifications from domain-blocked-but-followed people, do
not send Salmons to domain blocked users
* Add test
* Personal domain blocks shouldn't affect Salmon after all, since in this
direction of communication the control is very thin when it comes to
public stuff. Best stay consistent and not affect federation in this way
* Ignore followers and follow request from domain blocked folks,
ensure account domain blocks are not created for empty domain,
and avoid duplicates in validation
* Purge followers when blocking domain (without soft-blocks, since they
are useless here)
* Add tests, fix local timeline being empty when having any domain blocks
* feat(eslint): Set react/jsx-no-bind: error
* refactor(notifications/setting_toggle): Do not use bind
* refactor(components/dropdown_menu): Do not use bind
* refactor(components/autosuggest_textarea): Do not use bind
* refactor(compose/privacy_dropdown): Do not use bind
* refactor(compose/upload_form): Do not use bind
* refactor(components/status): Do not use bind
* refactor(components/onboarding_modal): Do not use bind
* refactor: PR feedback
* chore(notifications/setting_toggle): Lint
* refactor: PR feedback
The `params` variable here was quite overloaded.
It exists via the controller to hold the request params, and was sometimes being
used in this helper as that object, but other times was being used as a local
variable, or to pass to another method, and this was confusing.
This change renames the args for a method away from `params` for more clarity,
and extracts the actual usage of the controller-provided `params` to a
better-named method for clarity.
* Add <ostatus:conversation /> tag to Atom input/output
Only uses ref attribute (not href) because href would be
the alternate link that's always included also.
Creates new conversation for every non-reply status. Carries
over conversation for every reply. Keeps remote URIs verbatim,
generates local URIs on the fly like the rest of them.
* Conversation muting - prevents notifications that reference a conversation
(including replies, favourites, reblogs) from being created. API endpoints
/api/v1/statuses/:id/mute and /api/v1/statuses/:id/unmute
Currently no way to tell when a status/conversation is muted, so the web UI
only has a "disable notifications" button, doesn't work as a toggle
* Display "Dismiss notifications" on all statuses in notifications column, not just own
* Add "muted" as a boolean attribute on statuses JSON
For now always false on contained reblogs, since it's only relevant for
statuses returned from the notifications endpoint, which are not nested
Remove "Disable notifications" from detailed status view, since it's
only relevant in the notifications column
* Up max class length
* Remove pending test for conversation mute
* Add tests, clean up
* Rename to "mute conversation" and "unmute conversation"
* Raise validation error when trying to mute/unmute status without conversation
* Adding account domain blocks that filter notifications and public timelines
* Add tests for domain blocks in notifications, public timelines
Filter reblogs of blocked domains from home
* Add API for listing and creating account domain blocks
* API for creating/deleting domain blocks, tests for Status#ancestors
and Status#descendants, filter domain blocks from them
* Filter domains in streaming API
* Update account_domain_block_spec.rb
- add translation "Bon Appetoot!"
- add whitelist what doesn't require translation
- replace exclamation mark full width
- add space after question mark
- remove blank between words
- run yarn manage:translations
* fix(polyfil): Catch error from promise
https://pouchdb.com/2015/05/18/we-have-a-problem-with-promises.html
* fix(catch): Error
Make import promise catch a noop with commented out console.log for a debugging hint.
* fix(eslint): Disable eslint console rule
* fix(application): Use console error
* Always set `overflow: auto` to allow scroll just after opening
* Remove bounce animation which may cause unintended behavior due to max-height
* Use CSS transition instead of react-motion
* Some CSS refactoring including className changing
* Create common chunk rather than vendor chunk
vendor chunk is a set of modules provided by external vendors, but now we
can have a chunk as a set of modules shared by multiple entry points,
which could be more efficent than having vendor chunk.
* Start rails-ujs in common.js
This is used by /settings/two_factor_authentication.
* Add <ostatus:conversation /> tag to Atom input/output
Only uses ref attribute (not href) because href would be
the alternate link that's always included also.
Creates new conversation for every non-reply status. Carries
over conversation for every reply. Keeps remote URIs verbatim,
generates local URIs on the fly like the rest of them.
* Conversation muting - prevents notifications that reference a conversation
(including replies, favourites, reblogs) from being created. API endpoints
/api/v1/statuses/:id/mute and /api/v1/statuses/:id/unmute
Currently no way to tell when a status/conversation is muted, so the web UI
only has a "disable notifications" button, doesn't work as a toggle
* Display "Dismiss notifications" on all statuses in notifications column, not just own
* Add "muted" as a boolean attribute on statuses JSON
For now always false on contained reblogs, since it's only relevant for
statuses returned from the notifications endpoint, which are not nested
Remove "Disable notifications" from detailed status view, since it's
only relevant in the notifications column
* Up max class length
* Remove pending test for conversation mute
* Add tests, clean up
* Rename to "mute conversation" and "unmute conversation"
* Raise validation error when trying to mute/unmute status without conversation
From #2327 - Elephant friend was overlapping with text, oversized in
single column layout
From #2021 - Centered layout goes against design principles, changes
UX for everybody who's already used to the current one
From #2271 - CPU/RAM overusage from keeping columns in DOM (fix#2648,
possibly also #2101)
* Do not cancel PuSH subscriptions after encountering "permanent" error response
After talking with MMN about it, turns out some servers/php setups do
return 4xx errors while rebooting, so this anti-feature that was meant
to take load off of the hub is doing more harm than good in terms of
breaking subscriptions
* Update delivery_worker.rb
* Fix color contrast some more in privacy warning
Latest master appears to have changed the <strong> to this unreadable grey color. If you want that to be white then it should be pure white. If someone would rather revert this to that strong dark blue color it was before that would be good too.
* Make domain dark blue again
* add missing ;
woops
* Add <ostatus:conversation /> tag to Atom input/output
Only uses ref attribute (not href) because href would be
the alternate link that's always included also.
Creates new conversation for every non-reply status. Carries
over conversation for every reply. Keeps remote URIs verbatim,
generates local URIs on the fly like the rest of them.
* Fix conversation migration
* More spec coverage for status before_create
* Prevent n+1 query when generating Atom with the new conversations
* Improve code style
* Remove redundant local variable
The current text contrast on the privacy warning is a WCAG violation. I didn't notice this because my instance has a custom theme which is better. On default theme I am barely able to read the text with my impaired vision. This patch brings the contrast to Normal Text WCAG AA compliance, and Large Text WCAG AAA compliance.
Use ES Class Fields & Static Properties (currently stage 2) for improve class outlook.
Added babel-plugin-transform-class-properties as a Babel plugin.
* Fix#2027 - Accept own ID for remote follow with and without preceding @
Fix#2177 - Omit leading "acct:" in remote follow redirect template expansion
* Fix test
Receiving instances will then use their own missing image
Also, add <content /> to deleted statuses, since there was a reported
problem with the deletes and GNU social
* Fix regressions from #2683
Properly format spoiler text HTML, while keeping old logic for blankness intact
Process hashtags and mentions in spoiler text
Format spoiler text for Atom
Change "show more" toggle into a button instead of anchor
Fix style regression on dropdowns for detailed statuses
* Fix lint issue
* Convert spoiler text to plaintext in desktop notifications
* services: scan spoiler_text for hashtags (#699)
* views: link hashtags from spoiler_texts
This covers linking hashtags from within the spoiler
text on the server-generated pages.
* services: fix string concat going into hashtag RE
Cleaner Ruby syntax, may handle immutable strings better
Compact Language Detector v3 (CLD3) is the successor of CLD2, which was
used in the previous implementation. CLD3 includes improvements since CLD2,
and supports newer compilers. On the other hand, it has additional
requirements and cld3-ruby, the FFI of CLD3 for Ruby, is still new and may
be still inmature.
Though CLD3 is named after CLD2, it is implemented with a neural network
model, different from the old implementation, which is based on a Naïve
Bayesian classifier.
CLD3 supports newer compilers, such as GCC 6. CLD2 is not compatible with
GCC 6 because it assigns negative values to varibales typed unsigned.
(see internal/cld_generated_cjk_uni_prop_80.cc) The support for GCC 6 and
newer compilers are essential today, when some server operating system
such as Ubuntu Server 16.10 has GCC 6 by default.
On the one hand, CLD3 requires C++11 support. Environments with old
compilers such as Ubuntu Server 14.04 needs to update the system or install
a newer compiler.
CLD3 needs protocol buffers as a new dependency. However,it is not
considered problematic because major server operating systems, CentOS and
Ubuntu Server provide them.
The FFI cld3-ruby was written by me (Akihiko Odaki) for use in Mastodon.
It is still new and may be inmature, but confirmed to pass existing tests.
ExtractTextWebpackPlugin extracts the content of loaded files, which means
it loads files loaded by require.context but not required after that.
However the former implementation is not aware of that.
require.context can have a RegExp to filter files to load. This change
gives a RegExp which matches with SCSSes with 'custom' prefix to
require.context to take advantage of the feature.
