Saves HTML text nodes and selected element attributes in `htmltotext.txt` for each Snapshot. Primarily intended to be used for search indexing.