tl icon indicating copy to clipboard operation
tl copied to clipboard

Address overflow in `find_last_node_handle`

Open amandasystems opened this issue 1 year ago • 0 comments

I had some problems with find_last_node_handle() smashing its stack. This PR replaces the recursive version with an iteration that should have similar performance characteristics and which can safely parse inputs that the previous version could not.

I wish I had a copy of the document that caused the issue; unfortunately the crash was deep enough in processing the CommonCrawl dataset that it was difficult to find the HTML.

amandasystems avatar Feb 06 '24 15:02 amandasystems