Felix Boehm
Felix Boehm
> Do I in fact have to set xmlMode: false even in that case? You don't!
Intentional; if you parse XML, you will get self-closing tags. The slim export can work around this, as can disabling xml mode.
Assuming everyone has worked around this by now — sorry for the added work!
`.root()` returns a document node, which contains all other nodes. This is the same for every kind of markup. I've added support for `.prop` calls on document nodes in #4320,...
These types can now be imported from the domhandler module directly.
@dependabot rebase
Thanks for the report! Our CI still passes on Node 18, but I've bumped the minimum engines Node version to line up with Undici.
The decode data is packed this way for performance & memory usage (see https://github.com/fb55/entities/blob/main/scripts/trie/README.md). Unfortunately there is no way to fix this without degrading these for other users.
Thanks a lot @kainstar! Could you also add a test case? (The example you gave in the PR description would suffice.)
An HTML parser is operating on a token stream, one character at a time, so embedding a trie is quite natural. The trie is a considerable speed-up of the parser,...