import epitran

def to_lroh(s):
    s = s.replace('ɖ', 'ḍ')
    s = s.replace('ɾ', 'r')
    s = s.replace('ɽ', 'ṛ')
    s = s.replace('ʃ', 'š')
    s = s.replace('ʈ', 'ṭ')
    s = s.replace('j', 'y')
    s = s.replace('d͡ʒ', 'j')
    
    
    
    s = s.replace('ɑ̃ː', 'ɑ̃ɑ')
    s = s.replace('ɑː', 'ɑɑ')
    s = s.replace('ẽː', 'eẽ')
    s = s.replace('eː', 'ee')
    s = s.replace('ĩː', 'iĩ')
    s = s.replace('iː', 'ii')
    s = s.replace('ɔ̃ː', 'ɔ̃ɔ')
    s = s.replace('ɔː', 'ɔɔ')
    s = s.replace('ũː', 'uũ')
    s = s.replace('uː', 'uu')
    
    s = s.replace('ɑ', 'a')
    s = s.replace('̃ɑ', 'ã')
    
    s = s.replace('ɔ̃', 'õ')
    s = s.replace('ɔ', 'o')
    
    return s

def to_roheng(s):
    s = s.replace('ɖ', 'dh')
    s = s.replace('ɾ', 'r')
    s = s.replace('ɽ', 'ç')
    s = s.replace('ʃ', 'c')
    s = s.replace('ʈ', 'th')
    s = s.replace('j', 'y')
    s = s.replace('d͡ʒ', 'j')
    
    s = s.replace('ɑ', 'a')
    s = s.replace('̃ɑ', 'ã')
    
    s = s.replace('ɑ̃ː', 'ɑ̃ɑ')
    s = s.replace('ɑː', 'ɑɑ')
    s = s.replace('ẽː', 'eẽ')
    s = s.replace('eː', 'ee')
    s = s.replace('ĩː', 'iĩ')
    s = s.replace('iː', 'ii')
    s = s.replace('ɔ̃ː', 'ɔ̃ɔ')
    s = s.replace('ɔː', 'ɔɔ')
    s = s.replace('ũː', 'uũ')
    s = s.replace('uː', 'uu')
    
    s = s.replace('ɑ', 'a')
    s = s.replace('̃ɑ', 'ã')
    
    s = s.replace('ɔ̃', 'õ')
    s = s.replace('ɔ', 'o')
    
    return s

def convert_script(input_script, output_script, input_text):
    epi = epitran.Epitran(input_script)

    # store indices for capitalized words (will assume only first letter is capitalized)
    words = input_text.split()
    capital_indices = [i for i, word in enumerate(words) if word[0].isupper()]

    grapheme_text = epi.transliterate(input_text)

    if output_script == 'rhg-roheng':
        inter_text = to_roheng(grapheme_text)
    elif output_script == 'rhg-lroh':
        inter_text = to_lroh(grapheme_text)
    
    # reapply capitalization
    words = inter_text.split()
    for i in capital_indices:
        if i < len(words):
            words[i] = words[i].capitalize()
    output_text = ' '.join(words)

    return output_text

# print (f'Number of script mismatches: {numScriptMismatch} / {numEntries}')

# issues
# 
# ou
# glides with only one vowel nasalized (i.e is the whole glide always nasalized) (.e.g thiañ/ṭĩya) - need a constant way to deal with glides and nasalization (i.e. which vowel is nasalized?)