In current repository there will be functions to preprocessing text dataset and simple regex to create labels for dataset.
Labels - quotes, word with uppercase, Abbrev., cuts.
rt_reviews.csv - test dataset with sentences (not preprocessing!) if you want to check workability of script
Script read csv dataset and create another json with full sentence and list of labels from this dataset.