* Add a limit to how many posts can get fetched as a result of a single request
* Add tests
* Always pass `request_id` when processing `Announce` activities
---------
Co-authored-by: nametoolong <nametoolong@users.noreply.github.com>
* Return specific error on failure to parse Date header
* Add error message when preferredUsername is not set
* Change error report to be JSON and include more details
* Change error report to differentiate unknown account and failed refresh
* Add tests
* fix(status): remove send usage for private unlink_from_conversations
- make unlink_from_conversations public method
- rename unlink_from_conversations to unlink_from_conversations!
- fix send call on private method in statuses_vacuum and batched_remove_status_service
* fix(feeds_vacuum): replace find_in_batches with in_batches
because active record query results should be a little more efficient than
itterating with map and each. Postgres can grasp such lists of ids much quicker
than ruby can.
Will probably make allmost no difference, but cannot hurt either.
* Fix possible race conditions when suspending/unsuspending accounts
* Fix tests
Tests were assuming SuspensionWorker and UnsuspensionWorker would do the
suspending/unsuspending themselves, but this has changed.
There is an idempotency key generated by clients when authoring a post,
and stored in Redis, to ensure that if a user or client retries posting
the same status, we don't get a duplicate.
Hachyderm.io has been experiencing some filesystem and database
performance issues, causing database writes to be slow. This can mean
that there are successful posts, but the reverse proxy returns 504
Gateway Timeout before the idempotency status has been updated; users or
clients who retry (such as Tusky which retries automatically, see
tuskyapp/Tusky#2951) can re-try the same post with the same idempotency
key before it has actually been recorded in Redis, leading to duplicate
posts.
To address this issue, move all of the database updates after the
initial transaction that creates the status into the
`postprocess_status!` method, so we can insert the idempotency key
immediately after the status has been created, significantly reducing
the window in which the status could be created but the idempotency key
not yet stored.
Note: this has not yet been tested; I'm submitting this PR for
discussion and to offer to the Hachyderm.io admins to try out to fix the
multiple posting problem.
Co-authored-by: Brian Campbell <brcampbell@beta.team>
* Fix trying to fetch posts from other users when fetching featured posts
* Rate-limit discovery of new subdomains
* Put a limit on recursively discovering new accounts
Before this change, the following error would cause VerifyAccountLinksWorker to fail:
NoMethodError: undefined method `downcase' for nil:NilClass
[PROJECT_ROOT]/app/services/verify_link_service.rb:31 :in `block in link_back_present?`
- Change verification to happen in `default` queue
- Change verification worker to only be queued if there's something to do
- Add `link` tags from metadata fields to page header of profiles
* Fix followers count not being updated when migrating follows
Fixes#19900
* Fix validation error in SynchronizeFeaturedTagsCollectionWorker
Also saves remote user's chosen case for hashtags
* Limit remote featured tags before validation
- Only exact search matches for queries with < 5 characters
- Do not support queries with `offset` (pagination)
- Return HTTP 401 on truthy `resolve` instead of overriding to false
* Change featured tag updates to add/remove activity
* Fix to check for the existence of feature tag
* Rename service and worker
* Merge AddHashtagSerializer with AddSerializer
* Undo removal of sidekiq_options
* Add LIMIT of featured tag to instance API response
* Add featured_tags_collection_url to Account
* Add synchronization of remote featured tags
* Deliver update activity when updating featured tag
* Remove featured_tags_collection_url
* Revert "Add featured_tags_collection_url to Account"
This reverts commit cff349fc27.
* Add hashtag sync from featured collections
* Fix tag name normalize
* Add target option to fetch featured collection
* Refactor fetch_featured_tags_collection_service
* Add LIMIT of featured tag to v1/instance API response
* Move ActivityPub::FetchRemoteAccountService to ActivityPub::FetchRemoteActorService
ActivityPub::FetchRemoteAccountService is kept as a wrapper for when the actor is
specifically required to be an Account
* Refactor SignatureVerification to allow non-Account actors
* fixup! Move ActivityPub::FetchRemoteAccountService to ActivityPub::FetchRemoteActorService
* Refactor ActivityPub::FetchRemoteKeyService to potentially return non-Account actors
* Refactor inbound ActivityPub payload processing to accept non-Account actors
* Refactor inbound ActivityPub processing to accept activities relayed through non-Account
* Refactor how Account key URIs are built
* Refactor Request and drop unused key_id_format parameter
* Rename ActivityPub::Dereferencer `signature_account` to `signature_actor`
* Add a more descriptive PrivateNetworkAddressError exception class
* Remove unnecessary exception class to rescue clause
* Remove unnecessary include to JsonLdHelper
* Give more neutral error message when too many webfinger redirects
* Remove unnecessary guard condition
* Rework how “ActivityPub::FetchRemoteAccountService” handles errors
Add “suppress_errors” keyword argument to avoid raising errors in
ActivityPub::FetchRemoteAccountService#call (default/previous behavior).
* Rework how “ActivityPub::FetchRemoteKeyService” handles errors
Add “suppress_errors” keyword argument to avoid raising errors in
ActivityPub::FetchRemoteKeyService#call (default/previous behavior).
* Fix Webfinger::RedirectError not being a subclass of Webfinger::Error
* Add suppress_errors option to ResolveAccountService
Defaults to true (to preserve previous behavior). If set to false,
errors will be raised instead of caught, allowing the caller to be
informed of what went wrong.
* Return more precise error when failing to fetch account signing AP payloads
* Add tests
* Fixes
* Refactor error handling a bit
* Fix various issues
* Add specific error when provided Digest is not 256 bits of base64-encoded data
* Please CodeClimate
* Improve webfinger error reporting
* Change unapproved and unconfirmed account to not be accessible in the REST API
* Change Account#searchable? to reject unconfirmed and unapproved users
* Disable search for unapproved and unconfirmed users in Account.search_for
* Disable search for unapproved and unconfirmed users in Account.advanced_search_for
* Remove unconfirmed and unapproved accounts from Account.searchable scope
* Prevent mentions to unapproved/unconfirmed accounts
* Fix some old tests for Account.advanced_search_for
* Add some Account.advanced_search_for tests for existing behaviors
* Add some tests for Account.search_for
* Add Account.advanced_search_for tests unconfirmed and unapproved accounts
* Add Account.searchable tests
* Fix Account.without_unapproved scope potentially messing with previously-applied scopes
* Allow lookup of unconfirmed/unapproved accounts through /api/v1/accounts/lookup
This is so that the API can still be used to check whether an username is free
to use.
* Change e-mail notifications to only be sent when recipient is offline
Change the default for follow and mention notifications back on
* Add preference to always send e-mail notifications
* Change wording
* Change post text edit to not be considered significant if it's identical after reformatting
* We don't need to clear previous change information anymore
* Require status edits to be explicit, except for poll tallies
* Fix tests
* Add some tests
* Add poll-related tests
* Add HTML-formatting related tests
Not completely sure this could actually have any ill effect, but if
`RemoveStatusService` fails to acquire a lock in an
`ActivityPub::ProcessingWorker` job processing a `Delete`, the status
is currently discarded and causes a job failure but the next time the
job is attempted, it will skip deleting the status due to it being
discarded.
This commit makes the behavior of `RemoveStatusService` a bit more consistent
in case of failure to acquire the lock.
* Optimize and clean up `response_to_recipient?` CTE
Marginally improve performances, and make the CTE much more readable
* Limit max depth in `response_to_recipient?` CTE
* Fix edits with no actual changes being allowed locally
* Fix edits with no actual changes being allowed through ActivityPub
* Fix false positive changes caused by description processing in model
* Fix not recording poll expiration update
* Fix test
* Revert changes to ProcessStatusUpdateService
* Various fixes and improvements
* Fix code style issues
* Various changes and improvements
* Add guard clause
