* shared model for chat and inbox
* disable inbox schema
* inbox: use separate model
* remove old code that used group.chat
* add back chat field (not used) and remove old tests
* remove inbox exclusions when loading user
* add GET /api/v3/inbox/messages
* add comment
* implement DELETE /inbox/messages/:messageid in v4
* implement GET /inbox/messages in v4 and update tests
* implement DELETE /api/v4/inbox/clear
* fix url
* fix doc
* update /export/inbox.html
* update other data exports
* add back messages in user schema
* add user.toJSONWithInbox
* add compativility until migration is done
* more compatibility
* fix tojson called twice
* add compatibility methods
* fix common tests
* fix v4 integration tests
* v3 get user -> with inbox
* start to fix tests
* fix v3 integration tests
* wip
* wip, client use new route
* update tests for members/send-private-message
* tests for get user in v4
* add tests for DELETE /inbox/messages/:messageId
* add tests for DELETE /inbox/clear in v4
* update docs
* fix tests
* initial migration
* fix migration
* fix migration
* migration fixes
* migrate api.enterCouponCode
* migrate api.castSpell
* migrate reset, reroll, rebirth
* add routes to v4 version
* fix tests
* fixes
* api.updateUser
* remove .only
* get user -> userLib
* refactor inbox.vue to work with new data model
* fix return message when messaging yourself
* wip fix bug with new conversation
* wip
* fix remaining ui issues
* move api.registerLocal, fixes
* keep only v3 version of GET /inbox/messages
* rename hasAccess to canJoin for challenges
This is so the function won't be used accidentally for other
purposes, since hasAccess could be misinterpretted.
* add isLeader function for challenges
* allow challenge leader to join/modify/end challenge when they're not in the private group it's in
* delete duplicate test
* clarify title of existing tests
* add tests and adjust existing tests to reduce privileges of test users
* fix lint errors
* remove pointless isLeader check (it's checked in canJoin)
* new inbox client
* add tests for sendPrivateMessage returning the message
* update DELETE user message tests
* port v3 GET-inbox_messages
* use v4 delete message route
* sendPrivateMessage: return sent message
* fix
* prevent quest progress message in party chat when user is resting in the inn
* improve comment
* update tests now that the test group includes a new member (sleeping quest participant)
* adjust a test to fix lint failure (and make the test better)
* fix order of element assignments in test array
* remove commented-out code for purging PMs - no longer needed
https://github.com/HabitRPG/habitica/issues/7940#issuecomment-406489506
* adjust comments
* move cron code when sleeping / resting back into main body of cron code
* rename tests to use consistent terminology for sleeping
* add tests for cron when user is sleeping
* move sleeping tests to same place as non-sleeping test
This matches how the code has sleeping and non-sleeping code mingled.
* replace a broken test with new tests
The deleted test wasn't working correctly. The check that the user's
health hadn't decreased would have worked even if the user wasn't
sleeping because the Daily had been marked completed.
The new tests test both no damage from incomplete Dailies and
that Dailies are reset.
* add tests for Perfect Day buff and rename existing tests for consistent terminology
* remove old test code
* Fixing healing light not being castable on server and client sides when user has already full health
Adding integration test for spell cast of healing light when full health
Adding test for heal cast if user has full health
* Fixing ESLint syntax in the spells test files
* Added mute end date
* Added indefinite mute for users using slurs
* Fixed user reload. Added no longer muted message. Added format for date
* Fixed lint
* remove duplicate module.exports statement
* remove commented-out footer in Slack slur notification
There's no need for anything to replace this footer.
* swap order of flag actions to put most critical first
This causes moderators to be notified before the flagged message's flagCount is incremented, because if something happens to prevent the flagGroupMessage Promise from resolving, we still want to mods to see the notification.
* limit chat message flagging ability for new players
Players who created accounts less than three days ago can flag posts
but that does not contribute to the posts' flagCount. This prevents
a troll from maliciously hiding innocent messages by creating new
accounts to flag them.
* add tests
* fix other tests
* WIP - prevent multiple notifications
* merge promises to one
* update test, iterate each user
* revert changes in `groups.js` - filter duplicate notifications in `convertNotificationsToSafeJson`
* Fix joinedChallenge achievement being awarded when creating a challenge
* Modify test to check that achievement is not awarded for creating a challenge
* WIP(groups): add shared completion prop
Also fix an issue where the Needs Approval toggle would not read/save
correctly.
* fix(groups): save group options on task create
Also, correct count of assigned members when viewing user is among
assignments
* fix(groups): display correct messages in two places
* fix(tasks): eliminate console error related to filtering
Also localize a group plans string
* WIP(groups): implement single completion for approval workflow
* WIP(groups): Add shared completion handling to no-approval-needed flow
* WIP(groups): cover approval flow case for all-assigned
Also save new field on initial task creation
* fix(tasks): use default sharedCompletion value when creating tasks
* WIP(tests): non-working draft test
* Added completed todo to group query
* WIP(group-tasks): fix bugs, add tests
* refactor(group-tasks): deleteMany op, add more tests
* refactor(group-tasks): move shared completion handling to lib
* WIP(group-tasks): broken refactor
* WIP(group-tasks): await all the things
* Turned complete master task to save
* WIP(group-tasks): show completed
* fix(filtering): don't try to filter if no list is passed
* refactor(group-tasks): load completed to-dos on demand, not at start
* fix(group-tasks): don't double up on repeat visits
* fix(group-tasks): include brief explanation in dropdown
* fix(group-tasks): improve wording some more
* fixing typos in comments. yes, I am that kind of nerd
* replacing push-notify with node-apn in deps and in pushNotifications.js
* updating calling code and tests to use node-apn
* updating APN configs to new format
* migrating team ID and key ID to config.json
* update code to use env variables and add correct topic
* 10282: Added code for blocking party and guild invitations from blocked players, added tests
* 10282: fixed test label
* Update POST-groups_invite.test.js
removed `it.only` which was used for testing
* added 1px margin-right to .member-stats
* added unit test for flicker prevention style
* remove .only() from unit test
* rewrote margin test using computed style
* initial refactor
* add scoredUp and scoredDown values for habits history entries, one entry per habit per day
* fix lint and add initial migration
* update old test
* remove scoreNotes
* dry run for migration
* migration fixes
* update migration and remove old test
* fix
* add challenges migration (read only)
* fix challenges migration
* handle custom day start
* update tasks in migration
* scoring: support cds
* add new test
* add timestamp to moderator Slack messages
* fix test errors
* import moment, condense formatting
* add timestamp to author_name variable
* update test to include timestamp, fix footer matching
* change ISODate to Date
* update test to include timestamp
* Fixes asynchronous cron bug that allows cron to run twice for a user within seconds of eachother.
fixes#8991
* Fixing tests.
* Updating assignment to keep user and res.locals.user in sync.
* Revert commenting out some cron subscription tests
This reverts commit 47c488967c.
We're going to properly fix these tests, so let's start by reverting the
commit that temporarily disabled these tests in the first place.
Signed-off-by: aszlig <aszlig@nix.build>
* tests/cron: Fix restoring clock on test failure
Ah, the joys of global state... >:-(
Whenever some test failed which has mocked the time using
useFakeTimers(), other test that are run after that test would fail (or
even time out) as well, which is a bit confusing to debug.
Some of the tests even had a cleanup routine in afterEach() but most of
them didn't, so I rearranged them in a way so that we have a clock
variable for *all* of the subtests, which initially is null and then a
cleanup handler (also for *all* of the subtest) calls clock.restore() if
the value isn't null.
In order to avoid calling clock.restore() twice, I have removed all the
clock.restore() calls at the end of the tests setting the clock to a
specific value.
Signed-off-by: aszlig <aszlig@nix.build>
* tests/cron: Fix test for 3-month gift subscription
So this is the actual culprit of the test failures that emerge during
the first two days of a month:
The test group "for a 3-month gift subscription (non-recurring)" creates
a User object available for every test case, which has a subscription
for 3 months beginning at the current time/date.
During each test case the fake timer is set to the second day of the
month to be tested. For the first and second month it's unproblematic
because the subscription is still active, no matter whether the
dateTerminated is set to the first day or the last day of a month.
However, the third month is problematic here, because whenever the
subscription lasts until the first day of the third month it has already
ended after the second day and thus the test fails because the actual
implementation of cron sets plan.consecutive.count to zero (which is
what it's supposed to do).
In order to fix this, I've set dateTerminated for the User object to the
15th of the current month so the subscription lasts long enough to not
trigger the test failure (and also make time zones irrelevant, because
right now there is no TZ offset which is more than half of a month,
especially not while running the test suite).
Signed-off-by: aszlig <aszlig@nix.build>
