my company is interested in converting mhtml to html with embedded images as base64 in it. all in memory without saving to disk on the background. If I try to use your "document.save" API and save to stream, it is required to specify a location on disk to save images. we are not interested in that. your "save" code is not open source,so we can't see how the conversion is actually made. Is there a utility you providethat takes mhtml stream/text content, and retrieves html as stream/text with images embedded?