Inputs: dictionary.csv, public.csv, HakkaWords.csv, WaitauWords.csv
Process: compile.py
Outputs: chars.csv, hakka_words.csv, waitau_words.csv
In addition to words from HakkaWords.csv and WaitauWords.csv, extra words are automatically generated from collocations from the note column of dictionary.csv and entries with frequencies ≥ 10 from public.csv. Only entries which include at least one polyphone in the target language are included.