Fix several Problems with Provider Configuration

[ricoberger]

This commit contains several small fixes with the configuration of the cluster providers, the added fixes are listed below:

• Use global list for AWS regions: Until now the supported AWS regions where maintained in across several widgets. Now we are using a globally available variable "awsRegions" to maintain them, so that we only have to add new regions at one place.
• The json parsing of the providers when they were read from disk contained a bug, so that the providers doesn't contain the provider type after they were read. This is now fixed, so that the provider type is also available after the providers were read from our storage.
• The created clusters for the AWS and Google provider didn't contain the "default" namespace, so that automatically all namespaces where selected. This is now changed, so that the "default" namespace is used like for the other providers.
• The AWS SSO handling was improved, by disabling the buttons which can not be used until a required action was executed. This should help users to perform all actions in the correct order.
• The refresh token handling for the Google provider was fixed. Until now it was only possible to use the OAuth client one time, because the seconde time the provider was used the response from Google didn't contain a refresh token, so that we were not able to use the provider. To fix this issue we had to add the "prompt=consent" parameter to the login url.