This HTML parser reads HTML from an INPUT_STREAM and builds some HTML_ELEMENT objects known collectively as the roots. A correctly built HTML file should have only one root but this parser is more laxist.
ANY