* MM-12083: unittest using externally managed database
* cherry-pick Makefile changes from @cpanato
* Jenkins changes (#9915)
* add docker compose
* udpate
* when using minio dont need to set the region
* update
* add wait for it script
* using old minio
* add new jenkins file
* update makefile
* add dockerfile
* rename the docker-compose proj to avoid colision (#9917)
* rename the docker-compose proj to avoid colision
* enable debug
* enable debug to double checkt the branchs and fix docker-compose name (#9919)
* add ee hash to check (#9920)
* fix name (#9921)
* update jenkins file to push from branch and prs
* if a new push comes in stop the running build
* split mysql and postgres variables
* add script to run jenkins-like env in local dev env
* update docker-compose project name to use uuid to make it more randon
* fix DCNAME definition
* update elasticsearch docker image
* revert test
* tidy up stages, and wait for mysql differently
* update docker image and add check for postgres
* checking if is ready
* update docker compose to have a wait for deps
* add readme and rename dockerfile
* fix -unittest setup
* using mm docker image
* restore parallel unit tests at the package level
Spin up a dedicated database for each package under test to avoid races in accessing the same tables.
Simplify the interface for configuring the test database to just a DSN instead of multiple exports for each field.
* try to work around root mysql access in CI
* update local-test-env.sh too
* MYSQL_ROOT_HOST: %
* fix missing quotes
* setting some memory limits for mysql
* revert memory docker compose does not support
* fix env name for postgres
* expose errors in app/export_test.go
* fix test label, better error checking on teardown
* increase query timeout for tests
* fix export_test
* update local dev script
* add configurable mysql root passwd
* testlib: introduce and leverage
This doesn't yet factor out the individual test helpers: many packages
still rely on `api4` directly to do this, but now wire up the test store
setup through this package. `app` and `store`, in particular, don't use
`testlib` because of circular dependencies at the moment.
* cmd: command_test.go: use api4 testlib
* cmd: plugin_test.go: remove dependence on test-config.json
* cmd: config_test.go use configured database settings
* ensure test-(te|ee) exit with status code
* test-server: run all tests, deprecating test-te/test-ee
* cmd/mattermost/commands: fix unit tests
Instead of relying on (and modifying) a config.json found in the current path, explicitly create a temporary one from defaults for each test. This was likely the source of various bugs over time, but specifically allows us to override the SqlSettings to point at the configured test database for all tests simultaneously.
* wrap run/check into a test helper
It was insufficient to set a config for each invocation of CheckCommand or RunCommand: some tests relied on the config having changed in a subsequent assertion. Instead, create a new test helper embedding api4.TestHelper. This has the nice advantage of cleaning up all the teardown.
* additional TestConfigGet granularity
* customized config path to avoid default location
* be explicit if the storetest initialization fails
* generate safe coverprofile names in the presence of subtests
* additional TestConfigShow granularity
* fix permission_test.go typo
* fix webhook tests
* actually flag.Parse() to skip database setup on os.Execed tests
* fix recent regression in #9962, not caught by unit tests
* Moving goroutine pool
* Auto refactor
* Moving plugins.
* Auto refactor
* Moving fields to server
* Auto refactor
* Removing siteurl duplication.
* Moving reset of app fields
* Auto refactor
* Formatting
* Moving niling of Server to after last use
* Fixing unit tests.
Config Checks at StartUp Part1
Config Checks; Tests for TLS Server
HSTS header implementation + tests
make gofmt happy with new go version...
make gofmt happy with new go version #2...
fix logic bug
fix typo
Fix unnecessary code block
* #MM-12130 changes for custom service terms
* Fixed styling
* Added getServiceTerms API
* removed unnecessary panic
* removed custom service terms text from flat config
* reverted user sql store as those changes are no longer needed
* added tests
* Updated a config key to be more standard
* Added copyright info
* Loading service terms only if the feature is enabled
* Loading service terms only if the feature is enabled
* removed unused index
* added createservice termns API
* made a param to bool instead of string
* added createservice termns API
* review fixes
* fixed styling
* Minor refactoring
* removed saveConfig and loadConfig magic
* added empty service terms text check to createServiceTerms API
* refactoed some urls to be terms_of_service instead of service_terms
* removed check for support settings
* changed URLs in tests
* removed unused code
* fixed a bug
* added service termd id in conif
* fixed a test
* review fixes
* minor fixes
* Fixed TestCreateServiceTerms
* plugins: compute bundle hash on load
Use this hash to bust client caches whenever the plugin bundle changes.
* eliminate redundant pluginHandler
* switch to 64-bit FNV-1a
* Fix test
* Add ucLive support
crazy testing
lovely logs
more cookie work arounds
Added Access-Control-Expose-Headers to user login
Add complete_saml_body template and revert loginWithSaml endpoint
Set Access-Control-Allow-Credentials to true in user login
Login via email instead of username
Clean up code
Add comment to give some context
Move faml logic into saml function
Communicate via chrome sendMessage api
Remove unused code
Add config to support multiple extensions
Clean up embedded complete_saml template
Fix indentation for templates
Added license header to extension.go
Add EnableExperimentalExtensions flag
Extension validated for email auth
Clean up api auth code
Remove complete_saml_body.html
* Add extension support in saml
* Clean up code
* Clean up extension validation
* fix context's SiteURLHeader to use the subpath
This fixes various commands such as /open and /groupmsg, as well as
OAuth related redirects.
* MM-11246: also leverage GetSiteHeader() in completeSaml
* Adding blank request context to plugin hooks for future use.
* Rename RequestContext to Context
* Adding context to ServeHTTP and ExecuteCommand
* Fixing import cycle in test.
* factor out GetSubpathFromConfig
* mv web/subpath.go to utils/subpath.go
* serve up web, api and ws on /subpath if configured
* pass config to utils.RenderWeb(App)?Error
This allows the methods to extract the configured subpath and redirect
to the appropriate `/subpath/error` handler.
* ensure GetSubpathFromConfig returns trailing slashes deterministically
* fix error 404 handling
* redirect /subpath to /subpath/
This is necessary for the static handler to match, otherwise none of the
registered routes find anything. This also makes it no longer necessary
to add trailing slashes in the root router.
Examine ServiceSettings.SiteURL on startup and rewrite assets
accordingly if not in a development environment.
Also export `mattermost config subpath` command to manually do same.
This accompanies a webapp PR to use the updated `root.html` to define
the necessary webpack asset path for dynamically loading assets.
* Remove api package
* Remove api dependency from cmd package
* Remove EnableAPIv3 setting
* Update web tests
* Add more websocket tests
* Move some ws and oauth tests to api4 package
* Move command tests into api4 package
* Test fixes
* Fix msg command test
* Add some app file tests
* Refactor context out of API packages
* Update function names per feedback
* Move webhook handlers to web and fix web tests
* Move more webhook tests out of api package
* Fix static handler
* Implementing structured logging
* Changes to en.json to allow refactor to run.
* Fixing global logger
* Structured logger initalization.
* Add caller.
* Do some log redirection.
* Auto refactor
* Cleaning up l4g reference and removing dependancy.
* Removing junk.
* Copyright headers.
* Fixing tests
* Revert "Changes to en.json to allow refactor to run."
This reverts commit fd8249e99b.
* Fixing some auto refactor strangeness and typo.
* Making keys more human readable.
* failing to find i18n shouldn't segfault
The server was trying to handle the fact that it couldn't find the i18n
directory, by emitting a translated log message...
* fix utils.FindDir
The attempts to find the directory in the parent or grandparent directory
don't work if the current working directory was inside `enterprise`, with
`enterprise` itself being a symlink as per the usual developer setup.
Recurse to the root of the filesystem, cleaning the path along the way
to work around this limitation (and allow tests to be run from an
arbitrarily deep nesting level.)
Fix corresponding usages to employ filepath.Join.
* failing to find html templates shouldn't segfault
* fail fast if the test user cannot be created
* rework utils.FindDir to retain backwards compatibility
* Add fix to not run browser check on Mobile Browsers or the App
* remove non safari and IE checks
* Replace useragent checking dependency and update UA tests
* change some wording
* change dependancy again to one with compatible licence
* Fix typos and clarify wording
* fix typo and comvert tests to use array
