extracts entites tagged with "<span class=" from html files
scans directory for all html files, extracts as csv files, adds "extracted" at the beginning of filenames.
examines body of html file, extracts entites only if they are tagged wtih class=, before extracting, marks plaint texts witth class="O" tag for outsider texts. This is a pre-processor for preparing y_pred array to be called from sklearn library functions, classification_report or f1_score
note: google gemini pro 1.5 created main logic of the program