Previously if the list was displayed above the cursor
position and the list was filtered to decrease in size
it would be displayed several rows above the cursor as
it shrinks.
Now the position is updated each time the list is populated
so it will always use the latest height when calculating
the position in the editor to place the select list at.
We can just namespace it by the extension name since this the tree view
will only have one directory view for a given directory. There's no need
to disambiguate further.
This prevents autosave from writing unmodified buffers to disk, which
was causing the tree view to rebuild directory contents based on the
write event just when switching tabs.
We no longer subscribe to `RootView` so we don't need to unsubscribe.
We can use `@subscribe` to subscribe to window, so we don't need to
genette a unique id for the editor in order to unsubscribe, because the
editor will unsubscribe automatically when removed now.
TextMatePackage is only designed to load resources out of a TextMate
bundle. It's used only at load time, and from that point out we only
refer to our own global `syntax` data structure to access the data that
it loads.
These changes are mostly aesthetic in nature. I just thought it would
be a good idea to have theme loading be parallel to package loading as
much as possible. So I localized more logic on the `atom` global.
We'll look first in the user themes directory, then in the built-in
themes directory. This allows us to default to 'IR_Black' and not crash
if the user doesn't setup their `~/.atom/themes` directory.
Undo all operations that were performed during an
autocomplete session and only retain the confirmed
operation.
Previously the undo stack would fill with all the
selections made before confirmation and the prefix
was restored to its original text when the operation
was cancelled.
There's a slight wrinkle in this commit… TextMate grammars sometimes store
the `foldStopMarker` directly in the grammar, rather than storing it
in a separate scoped preferences file like the other settings. So we
have to scan through grammars looking for those that have the fold end
marker and make a scoped property for that grammar's scope.
Previously, we had a custom method on the `TextMateBundle` class for
retrieving these variables from the bundle. Now we're using Atom's
`syntax.getProperty` mechanism. The idea is to map TextMate concepts
to their Atom equivalent, rather than building everything directly
around TextMate.
