Kibana XHR requests not encoded leading to Bad Request

[flowl]

Kibana version:
Elasticsearch version:
Server OS version:

7.5.1 deployed through https://cloud.elastic.co/

Browser version:

Most recent Chrome 27-01-2020 most recent OSX up today

Describe the bug:

I think either due to a change in fluentd or in ES, new index get named wrong,
instead of fluentd.postgres.20200115, everything ends up in fluentd.postgres.%y%m%d.

Due to this, I discovered that I cant look at the settings of the index with the% in the name. In the browser console the server responds with Bad Request due to the % not beeing encoded before the request.

The request is

/api/index_management/settings/fluentd.postgres.%y%m%d

instead of

/api/index_management/settings/fluentd.postgres.%25y%25m%25d

Steps to reproduce:

1. Create something with index name containing special characters like %.
2. Try to view the index' settings through Kibana

Screenshots (if relevant):

[elasticmachine]

Pinging @elastic/es-ui (Team:Elasticsearch UI)

[yuliacech]

Closing since the flyout has been replaced by an index details page in #165705