* Added webp package into app/emoji.go , also added a test
* fixed import order inside emoji.go
* Ran go mod tidy
* Conversion to png for Webp emojis that must be resized is implemented
* Deleted unused import
* Added reading mechanism for webp instead of creating
* fileutils and filepath are used while reading a webp file during test
* Deleted readtestwebp function and put it in emoji_test as requested
* Added webp image upload test inside api4/file_test.go
* Expect image is changed from false to true
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
* Add route endpoints, model, store functions, and tests for top threads
* Run make store-layers
* Make the following changes
- Fix top user threads query
- Fix passing parameters in api4/insights.go to handler in app
- Add top user threads test
* Add post-message, user_id, participants information to insights results
* model.TopThread.UserID -> model.TopThread.UserId, for compatibility with MySQL
* Rename name -> channel_name
* Add user information to response
* Link post in response, filter out deleted root posts from top threads
* Handle thread delete cases, add app tests for threads insights
* lint: fix typo
* lint: rename asserts
* lint: require.nil -> require.NoError
* Add integration tests for thread insights
* Add embeds and images to top posts
* Add license checks for top threads endpoints
* Query users in batch to populate post-creator
* Make the following changes
- Add license to test server in api4/
- Add tests for threads insights
- top team threads shouldn't include threads from other teams, DMs
- Test duration constraint
- Pagination testing for top threads in model/insights_test.go
* Add i18n-extract
* i18n fixes
* Add username, nickname to user_information
* Hide message, user_id, post_id, reply_count in depth=1 of top threads response
* Fix tests using response.reply_count to use response.post.reply_count
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
* [MM-44667] Set InsecureSkipVerify for S3 when EnableInsecureOutgoingConnections is set
* Cleanup s3New, customTransport
* Lint fixes
* Cleanup s3store_test.go
* Using default transport
```release-note
NONE
```
Co-authored-by: Jake Gutierrez <jakegut0108@gmail.com>
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
* if there is no trial end date, use the current date for the upgrade confirmation email
* changes
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
* Introduced search for channel id's
* Small fix to test case as it is not looking at count
* Introduced searching channels by id as a ChannelOpt. Defaulting to true for system console for the time being
* go fmt
* Modified whether to include search by id by passing it in as prop to search endpoint. Modified webapp to pass through search by id as well
* Fixed issue with full text search, as there is not always necessarily an index on Id.
* Modified test to verify count, and moved code inside conditional
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
Discard the non-essential WebSocket events before they are queued.
When conn send queue is full, the connection is closed. There are messages that are being dropped by the current mechanism when send queue is at 50% capacity. This change makes the messages drop before entering the queue in order to keep the queue as empty as possible.
The value of the length of the queue could not be 100% accurate since we are reading from the hub goroutine, but it will be accurate enough to avoid some events.
* MM-43956: Adds post counts by day.
* MM-43956: Test data cleanup. Switch from Unix to UnixMilli.
* MM-43956: Changes from selecting date data types to strings in SQL.
* MM-43956: Fixes date format key.
* MM-43956: Adds missing user id scope for 'my' top channels graph.
* MM-43956: Adds the ability to group post counts by hour.
* MM-43956: Require enterprise or professional license. Reject guests.
* MM-43956: Adds license for tests.
* MM-43956: Renames function.
* MM-43956: Omits future hours from post counts by hour.
* MM-43956: Adjust API response grouping to users timezone.
* MM-43956: Adds translation.
* MM-43956: Adds user's timezone to the data tier for the grouping by day and hour.
* MM-43956: Fixes layers.
* MM-43956: Lint fix.
* MM-43956: Fix store layers.
* MM-43956: Switches to default name for time package; changes parameter names to avoid naming conflict.
* MM-43956: Updates mocks.
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
```release-note
We create a new config option MaxImageDecoderConcurrency which
indicates how many images can be decoded concurrently at once.
The default is -1 which means number of CPUs present.
This affects the total memory consumption of the server. The
maximum memory of a single image is dictated by MaxImageResolution * 24 bytes.
Therefore, a good rule of thumb to follow is that MaxImageResolution
* MaxImageDecoderConcurrency * 24 should be less then the allocated memory for
image decoding.
```
https://mattermost.atlassian.net/browse/MM-43733
* MM-42581: fixes unread threads on user channel add
Currently when we are adding a user to a channel we don't send previous
values for unread replies and mentions. This is resulting the thread to
not be marked as unread in the UI, since we rely on the previous values
for that.
This commit fixes the issue by returning previous unread values of 0.
* Adds test
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
These logs would appear at every request
and didn't have any useful info for the admin.
It was more of a developer logging thing
and therefore removing it.
https://mattermost.atlassian.net/browse/MM-39690
```release-note
NONE
```
* MM-44412: correct usage of updateAt to createAt
* lint
* correct usage of updateAt to createAt - CRT unsupported
* remove dead code
* remove unrequired tests
* update test name
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
* MM-41728: Validate emoji size in app on import
* use NotNil instead of Error for AppErr
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
* store: guests will only receive results of which channels they are in
* api4/channel_test: add test case for guest accounts channel autocomplete
* apply to searchengine and also for AutocompleteInTeam
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
* Prevent cloud limited installations from restoring teams when at or above the teams limit
* Code clean up
* fix i18n
* Actually fix i18n
* updates for govet
* Update model/client4.go
Co-authored-by: Vishal <vish9812@gmail.com>
* [MM-44397] Restrict team creation based on subscription limits (#20282)
* restrict team creation based on limits
* Update api4/team.go
Co-authored-by: Maria A Nunez <maria.nunez@mattermost.com>
* Fix tests
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
Co-authored-by: Maria A Nunez <maria.nunez@mattermost.com>
* Add additional field to teamsusage
* Fix
* remove useless test
* Fix error for team creation
* Remove apostrophe
Co-authored-by: Mattermod <mattermod@users.noreply.github.com>
Co-authored-by: Vishal <vish9812@gmail.com>
Co-authored-by: Maria A Nunez <maria.nunez@mattermost.com>
Summary
In a bulk import, if the createAt of a reply is older than the createAt of the parent post, then instead of erroring, we log a warning and set the createAt of the reply to that of the parent post.
I checked the exporting logic, there doesn't appear to be away for the reply createAt to be older than the parent createAt, as a side-effect of the exporting logic.
Ticket Link
https://mattermost.atlassian.net/browse/MM-43878
* MM-38982: fixes SVG uploading
Uploading SVG files was "almost" treated like an image in some places,
resulting in not showing the SVG on the client.
This is happening because when we fail to generate a mini preview for an
image, we prepend "invalid-" to the MimeType.
This commit adds a check to guard against SVG files in methods that make
no sense for SVGs.
* Reverts invalid-{mimetype} fix
* Uses fileInfo.IsSvg() where it can
