The fact that these methods only return local editors (i.e., editors for the local user while hosting a Code With Me session) is an implementation detail
This means we listen to changes in all documents, rather than just the changes in the documents for open local editors. And this means that we correctly update e.g. marks when a non-local editor changes a file that isn't open in a local editor.
Reviews all IntelliJ listeners to ensure that they only work with supported local editors. Editor creation was initialising IdeaVim for all editors, which meant that behaviour could leak into Code With Me guest editors. E.g. guest editors incorrectly drawing relative line numbers, or the host using the guest's last selected tab when switching to an alternate file.
This leads to a change in behaviour with some local editors. The editor creation listener will now check to see if the editor is local *and supported*. This means it can exclude single line editors, editors in database cells or dialogs, depending on the state of 'ideavimsupport' at creation time. The behaviour at creation time is now more correct, but if 'ideavimsupport' is modified, existing matching editors will not be initialised.
Fixes VIM-3274, fixes VIM-3275
Sometimes it's not a plugin error and may indicate that key is propagated for later handling by IDE
But what we know for sure - that for both cases we should reset command builder
Applying default values may lead to unexpected results, especially if we sometimes want to use the global state (IJ), and at other times, its clone for asynchronous processing (Fleet).
It unnecessarily binds mappingState to mode and thus to editor. And we want to simplify things and have a single MappingState instead of multiple of them
Rationale:
1. A much more experienced developer, whom I highly respect, suggested to empty VimStateMachineImpl constructor in his TODO comment.
2. I aim for VimStateMachine to be a data class rather than being a container for both data and complex logic.
3. From an architectural perspective, it is more correct. Editors do have state (or they may possess a single global state if the corresponding option is set), but a state does not own an editor.
This is done because of platform changes. Now the `lastUpdateAndCheckDumb` doesn't update an action that supposed to be updated on background.
The problem was detected with commentary tests. The test supposed to use the line comment in case the block comment is not available. However, the since the action was not updated, the presentation was not reset to false and the fallback to line action was not performed.
