* removes from emailverificationtoken spec
* updates winston
* Updates ValidationAndPasswordsReset
* Use local request in schemas
* Removes request in rest.spec
* Removes request from PushRouter0
* removes request from public API
* removes request from index.spec
* Removes request form parse.push spec
* removes request from ParseInstallation spec
* Removes from ParseHooks
* removes request from ParseGlobalConfig.spec
* Removes request from ParseAPI.spec.js
* removes request from LogsRouter
* removes in features
* Filters undefined headers instead of crashing
* Removes request from ParseUser spec
* Removes usage of request in ParseFile.spec.js
* Removes request from AuthAdapters.js
* removes request-promise from ParseGeoPoint.spec
* Removes request-promise from ParseQuery spec
* remove request-promise from UserPII
* removes request-promise from EnableExpressErrorHandler
* Updates RevocableSessionUpgrade spec
* Update RestQuery
* Removes read preferenceOptionM
* ensure we forward auth from URL
* use request in CloudCode.spec.js
* Removes request-promise from JobSchedule.spec
* Removes rp from VerifyUserPassword.spec.js
* Removes rp from PasswordPolicy spec
* Removes rp from ParsePolygon spec
* Removes rp from fullTextSearch spec
* Removes rp from PArseQuery.Aggregate
* Ensure we properly forward errors
* Removes request and request-promise
* Use the correct function when validating google auth tokens
httpsRequest.request expects the param postData and has no default value
or validation to check if it is missing before using it. As a result, an
error `TypeError: First argument must be a string or Buffer` is
thrown when an attempt is made to authenticate with Google.
A quick check on the LinkedIn, FB, and twitter authentication adapters
shows they are using httpsRequest.get for their validation calls.
* Correct google auth adapter tests
* Refactor all auth adapters to reduce duplications
* Adds mocking and proper testing for all auth adapters
* Proper testing of the google auth adapter
* noit
* WIP: Integrate JS SDK v2
- Removes backbone style callbacks
- Use Promise instead of Parse.Promise
* Fixes ParseObject and ParseRelation
* Updates Parse.Query with promises
* Alls tests should pass
* Ensure a fresh user is used for each test
* Use REST implementation to avoid side effects for username/email duplicates
* Uses js sdk v2
* Various improvements in test name / de-duplications
* Reverts to class by class deletion, introduced fast mode that just delete data for mongo
- Speeds up are incredible Executed 1695 of 1713 specs INCOMPLETE (18 PENDING) in 4 mins 19 secs.
* Adds documentation about the deleteEverything
* Removes need to use babel-register
- Adds watch to watch changes when running the test to regenerate
- Tests are now pure node 8
* Adds timing to helper.js
* Update contribution guide
* Adds inline sourcemaps generation to restore coverage
* nits
* Integrate auth adapter for Facebook accountkit login
* Also verify Facebook app id associated with account kit login
* Add appsecret_proof as extra graph request parameter
* Specific error message for Account kit and more test coverage
* One more test to cover when AppIds for Facebook account kit not configured properly
* chore(package): update jasmine to version 3.0.0
Closes#4547
* Fixes failing tests for jasmine 3.0
Starting 3.0, done(something) will fail
* Update tests so they dont leverage var, but let and const
With jasmine 3.0, the randomization engine was making the test fails because of the scope of `var`
* Remove randomizer
* Use same adapter for PG tests, drop table to ensure the tests dont side effect
* Adds flow types / Configuration interfaces
* Lets call it options
* Use a single interface to generate the configurations
* Translates options to definitions only if comments are set
* improves logic
* Moves objects around
* Fixes issue affecting logging of circular objects
* fixes undefined env
* Moves all defaults to defaults
* Adds back CLI defaults
* Restored defaults in commander.js
* Merge provided defaults and platform defaults
* Addresses visual nits
* Improves Config.js code
* Adds ability to pass the default value in trailing comments
* Load platform defaults from the definitions file
* proper default values on various options
* Adds ParseServer.start and server.start(options) as quick startup methods
* Moves creating liveQueryServer http into ParseServer.js
* removes dead code
* Adds tests to guarantee we can start a LQ Server from main module
* Fixes incorrect code regading liveQuery init port
* Start a http server for LQ if port is specified
* ensure we dont fail if config.port is not set
* Specify port
* ignore other path skipped in tests
* Adds test for custom middleware setting
* Refactors new Config into Config.get
- Hides AppCache from ParseServer.js, use Config.put which validates
* Extracts controller creation into Controllers/index.js
- This makes the ParseServer init way simpler
* Move serverURL inference into ParseServer
* review nits
* Moves all authentication providers to Adapter/Auth
* refactors specs
* Deprecates oauth option in favor of auth option
- Deprecates facebookAppIds option (in favor of auth.facebook.appIds)
- Adds warnings about the deprecated options
* nits
