* Move signature verification stoplight to the requests themselves
This avoids blocking messages from known keys for 5 minutes when only one fails…
* Put the stoplight on the actual client IP, not a potential reverse proxy
* Refactor signature verification a bit
* Rescue signature verification if recorded public key is invalid
Fixes#8822
* Always re-fetch AP signing key when HTTP Signature verification fails
But when the account is not marked as stale, avoid fetching collections and
media, and avoid webfinger round-trip.
* Apply stoplight to key/account update as well as initial key retrieval
* Downcase signed_headers string before building the signed string
The HTTP Signatures draft does not mandate the “headers” field to be downcased,
but mandates the header field names to be downcased in the signed string, which
means that prior to this patch, Mastodon could fail to process signatures from
some compliant clients. It also means that it would not actually check the
Digest of non-compliant clients that wouldn't use a lowercased Digest field
name.
Thankfully, I don't know of any such client.
* Revert "Remove dead code (#8919)"
This reverts commit a00ce8c92c.
* Restore time window checking, change it to 12 hours
By checking the Date header, we can prevent replaying old vulnerable
signatures. The focus is to prevent replaying old vulnerable requests
from software that has been fixed in the meantime, so a somewhat long
window should be fine and accounts for timezone misconfiguration.
* Escape users' URLs when formatting them
Fixes possible HTML injection
* Escape all string interpolations in Formatter class
Slightly improve performance by reducing class allocations
from repeated Formatter#encode calls
* Fix code style issues
`@signature_verification_failure_reason` is used in most places but`@signed_verification_failure_reason` appears in two places. Likely those errors are not returned.
* Fix follow/unfollow buttons on public profile
- Present non-logged users with web+mastodon:// URLs for remote accounts
- Present logged-in users with appropriate links (authorize_follows and
remote_unfollows) for remote accounts
* Do not cache rendered cards if user is logged in
Previously the default locale was set by Localized concern for controllers,
but it was not enforced for mailers.
config is enforced throughout the application and an appropriate place to
set the default locale.
UserTrackingConcern is circumvented by SessionsController#create
because it calls warden, which calls the User#update_tracked_fields!
method directly. Move returning user logic to that method.
* Add GET /api/v1/instance/peers API to reveal known domains
* Add GET /api/v1/instance/activity API
* Make new APIs disableable, exclude private statuses from activity stats
* Fix code style issue
* Fix week timestamps
* Add moderator role and add pundit policies for admin actions
* Add rake task for turning user into mod and revoking it again
* Fix handling of unauthorized exception
* Deliver new report e-mails to staff, not just admins
* Add promote/demote to admin UI, hide some actions conditionally
* Fix unused i18n
- For some reason, :if option on before_action did not work. It got
executed every time, returned false, and the action run anyway,
which led to the current_sign_in_at and sign_in_count being
updated on every request
- Return "do not filter" early in FeedManager#filter_from_home? if
the status is authored by receiver. Usually this method is not
called for own statuses at all, but it is called when Feed#get
uses the database
- Return early if #reload_stale_associations! has nothing to load
to save a database query with WHERE 1=0
Additionally, ActivityPub::FetchRemoteStatusService no longer parses
activities.
OStatus::Activity::Creation no longer delegates to ActivityPub because
the provided ActivityPub representations are not signed while OStatus
representations are.
* Return sensible HTTP status for ActivityPub inbox processing
* Return sensible HTTP status for salmon slap processing
* Return additional information to debug signature verification failures
If the signature could not be verified and the webfinger of the account
was last retrieved longer than the cache period, try re-resolving the
account and then attempting to verify the signature again
- Tries to avoid performing HTTP request if the keyId is an actor URI
- Likewise if the URI is a fragment URI on top of actor URI
- Resolves public key, returns owner if the owner links back to the key
* Add Request class with HTTP signature generator
Spec: https://tools.ietf.org/html/draft-cavage-http-signatures-06
* Add HTTP signature verification concern
* Add test for SignatureVerification concern
* Add basic test for Request class
* Make PuSH subscribe/unsubscribe requests use new Request class
Accidentally fix lease_seconds not being set and sent properly, and
change the new minimum subscription duration to 1 day
* Make all PuSH workers use new Request class
* Make Salmon sender use new Request class
* Make FetchLinkService use new Request class
* Make FetchAtomService use the new Request class
* Make Remotable use the new Request class
* Make ResolveRemoteAccountService use the new Request class
* Add more tests
* Allow +-30 seconds window for signed request to remain valid
* Disable time window validation for signed requests, restore 7 days
as PuSH subscription duration (which was previous default due to a bug)
* 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!
* 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
* 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
This PR fixes I18n.locale for rake middlewares. Mastodon uses Devise that depends on Warden.
Warden::Manager can be found in rake middleware. It is outside of the controller.
In the case of authentication failed, warden calls throw(:warden). At the time Warden::Manager
delegates request to failure_app to generate response and flash[:alert] after catching it.
Unfortunately, I18n.locale is already reset then because I18n.with_locale is enabled only
inside the controller. If we used I18n.locale=, Devise::FailureApp could get the current locale.
* Add routing specs for accounts followers and following actions
* Use more restful route naming for public account follow pages
Moves two actions:
- accounts#followers to accounts/follower_accounts#index
- accounts#following to accounts/following_accounts#index
Adds routing spec to ensure prior URLs are preserved.
* Add locale spec with failing locale plus region check
* Use a more accurate locale when supplied by browser headers
Previously we were using a matching option which would use the first locale
available which matched the locale portion, even if a region was specified.
This changes to first try to find an exact match, and then fall back to the
region, and then fall back to the default.
* Clean up default_locale method
* Use HTTP Accept-Language to detect locale
* Fix gem order to comply with codeclimate
* Sort gem to comply with rubocop
* I18n.default_locale fallback when there is no accept-language header
Setting of locale in controller extracted to Localized concern,
the doorkeeper authorized applications controller moved under
custom namespace with inclusion of Localized, which resolves the
"it sometimes appears in a different random language" bug
