Repository navigation
Conversation
Building the DOM is quadratic in nesting depth and raw HTML from a paste reaches it verbatim, so a megabyte of `<div>` was a quarter million levels deep and cost around 87 seconds of CPU per request — on a blocking thread the request timeout does not stop. Prose stays around a dozen levels, so cap the depth at 256 and answer with a bad request instead. The depth is measured by running html5ever's own tree builder, set up as ammonia sets it up, over a sink that keeps nothing but where each node hangs. Estimating it from the tags falls short of the parser at every turn: a close tag naming nothing that is open closes nothing, and neither does one in a comment, an attribute value or a `<textarea>`; `<div/>` opens a level; a misnested `<b>` is reopened inside every element that follows it. Each shortfall hands the sanitizer the very tree the limit guards against, and only the parser itself gets all of them right. The document goes in a kilobyte at a time and the parse is abandoned once it is too deep, so the check cannot be made to build that tree either. On a megabyte of prose it adds about a tenth to the render. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This branch has not been deployed
This file contains hidden or 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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Building the DOM is quadratic in nesting depth and raw HTML from a paste reaches it verbatim, so a megabyte of
<div>was a quarter million levels deep and cost around 87 seconds of CPU per request — on a blocking thread the request timeout does not stop. Prose stays around a dozen levels, so cap the depth at 256 and answer with a bad request instead.The depth is measured by running html5ever's own tree builder, set up as ammonia sets it up, over a sink that keeps nothing but where each node hangs. Estimating it from the tags falls short of the parser at every turn: a close tag naming nothing that is open closes nothing, and neither does one in a comment, an attribute value or a
<textarea>;<div/>opens a level; a misnested<b>is reopened inside every element that follows it. Each shortfall hands the sanitizer the very tree the limit guards against, and only the parser itself gets all of them right.The document goes in a kilobyte at a time and the parse is abandoned once it is too deep, so the check cannot be made to build that tree either. On a megabyte of prose it adds about a tenth to the render.