Closed
Description
At the moment, the language server tries to pre-filter the bean completion proposals, matching the name of the beans against the prefix that is in the editor (the AST node, to be precise). This seems to be case sensitive (which causes some proposals to now show up when the user typed everything lower case, but it would match against a bean name with one of those characters being uppercase).
But instead of taking upper- vs. lowercase things into account on the language server side, we should always return all the bean completion proposals and let the client side (the editor) decide which proposals to filter out and which not.