Document copy constructor?

[crazydef]

Hi,

After profiling, I've noticed that parsing HTML source when creating a new document is a huge performance hit (I can be parsing the same HTML code hundreds of times).

Basically my work flow is: Create an object (in our engine), attach a litehtml document to it (which involves creating/parsing/etc) and then update parts of the parsed HTML with data from that object every frame - or whenever updates arrive over the network.

Is there an easy way to parse the document once and attach a duplicate of the pre-parsed data structure that I can then modify accordingly instead of parsing over and over and over again?