Add HTB sentence csv and read functions

[cjer]

• Add sentence csv with fields: global_sent_id, spmrl_sent_id, ud_sent_id, text_from_ud, tokens, set, duplicate_global_sent_id, very_similar_global_sent_id
• Add a read_sentence_df function in readers.py
• Add info from this DataFrame to CONLLU output sentence metadata fields