avm1: Store childNodes on the node itself #13319
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| @@ -222,6 +222,8 @@ impl<'gc> Xml<'gc> { | |||
| } | |||
| } | |||
|
|
|||
| self.root().refresh_cached_child_nodes(activation).unwrap(); // :( | |||
There was a problem hiding this comment.
I don't think the AVM1 operations done by refresh_cached_child_nodes can throw an error?
adrian17
force-pushed
the
avm1-cache-child-nodes
branch
2 times, most recently
from
d0a7c8f to
cadeede
Compare
Dinnerbone
approved these changes
Oct 9, 2023
Dinnerbone
force-pushed
the
avm1-cache-child-nodes
branch
from
cadeede to
50da6d6
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
I didn't do benchmarks yet. I can also add a test later.
This should help #13130 .
In general, this should make repeated
.childNodesa free op, at the cost of making mutations (.appendChildetc) more expensive (O(n_children)* but only if.childNodeswas read at least once before).This is not just an optimization, but also a correctness fix, as the array mutations can be observed in AS.
(*which implies that appending N elements is
O(n^2). I confirmed that FP has the same quadratic time complexity, though without benchmarks I can't tell whether we'll turn out faster or slower)Unfortunately it's ugly due to current design. Ideally we'd want to refresh the array each time a mutation of a
XmlNode(xml/tree.rs) happens, but this refresh is an AVM1 operation that technically can throw an AVM1 error, which makes actually calling this fromtree.rsmuch harder.Any advice? 😅