What is one of the main complexities when using HTMLParser?

It isn't a streaming parser which causes processing of data to be much slower.

It isn't a streaming parser which keeps track of all tags, making it difficult to process data

It is a streaming parser which requires the class to keep track of starting and ending tags