Support transforms for each widget #753
Conversation
Philipp-M
force-pushed
the
transforms
branch
from
7da5c16
to
b97a9e9
Compare
| pub fn iter(self) -> impl DoubleEndedIterator<Item = ArenaRef<'a, Item>> + ExactSizeIterator { | ||
| self.children | ||
| .iter() | ||
| .map(move |child| child.arena_ref(self.id, self.parents_map.parents_map)) | ||
| } |
There was a problem hiding this comment.
As I understand, this is added for performance reasons. #752 is another way to address those performance issues, but I wonder if we couldn't also solve them (in the safe TreeArena version) by having children be a HashSet?
Then checking "X is child of Y" goes from O(N) to O(log(N)), which is still ridiculously inefficient, but good enough for our safe version.
There was a problem hiding this comment.
Yeah that's probably a good trade-off and at least avoids quadratic slowdown (which practically results in being completely "stuck" with big enough N).
But I think you mean a HashMap, i.e. something like children: HashMap<u64, TreeNode<Item>>?
There was a problem hiding this comment.
Ok, should this go into this PR? Or a separate PR, as I think it's somewhat orthogonal?
There was a problem hiding this comment.
We'll see after #752 merges. Probably a separate PR. Either way, I'd prefer if this PR didn't add an iter() method.
There was a problem hiding this comment.
Sure, as I noted here (and in zulip) that this commit is temporary (a reason why I tried to isolate these changes, so that I can easily remove it again).
Alright then I'm waiting for the merger of #752 and implement it on top.
Checkout this zulip thread for more info/discussion.
This is far from the final state, and at this point should more serve for discussion.
See the xilem
masonorhttp_catsexample, which I have abused to get an initial demo working.