A python script that converts normal english text into my own phonetic latin-based english writing system. All phonemes are represented by digraphs and the only diacritics are optional and used to inform about stress and places where digraphs are actually pronounced separately. Example text: érlij tuw bed and érlij tuw rajz mejks a maen hélthij,…

convert.py

#!/bin/python
from big_phoney import BigPhoney, preprocessors
from os import devnull
from tqdm import tqdm
from contextlib import redirect_stdout
from unicodedata import normalize

phonemeMap = {
    # Vowels
    "aa": "aa",
    "ae": "ae",
    "ah": "a",
    "ao": "o",
    "aw": "aw",
    "ay": "aj",
    "eh": "e",
    "er": "er",
    "ey": "ej",
    "ih": "i",
    "iy": "ij",
    "ow": "ow",
    "oy": "oj",
    "uh": "u",
    "uw": "uw",
    # Consonants
    "b": "b",
    "ch": "tsh",
    "d": "d",
    "dh": "dh",
    "f": "f",
    "g": "g",
    "hh": "h",
    "jh": "dzh",
    "k": "k",
    "l": "l",
    "m": "m",
    "n": "n",
    "ng": "ng",
    "p": "p",
    "r": "r",
    "s": "s",
    "sh": "sh",
    "t": "t",
    "th": "th",
    "v": "v",
    "w": "w",
    "y": "j",
    "z": "z",
    "zh": "zh"
}

vowels = "aeiou"

alphabet = "qwertyuiopasdfghjklzxcvbnm"
alphabet += alphabet.upper()
punctuation = ".?!"

enableDiaeresis = True
enableStressMark = True

digraphs = []
trigraphs = []

for phoneme in phonemeMap.values():
    match len(phoneme):
        case 2:
            digraphs.append(phoneme)
        case 3:
            trigraphs.append(phoneme)


outputText = ""

word = ""
capitalizeNext = True

bigPhoney = BigPhoney(preprocessors = [preprocessors.ReplaceAbbreviations])

with open(devnull, "w") as devnull:
    for char in tqdm(open("input.txt", "r").read()):
        if char in alphabet:
            word += char
            continue

        if word != "":
            phonemes = []
            syllableCount = 0

            phonizedWord = ""
            with redirect_stdout(devnull):
                phonizedWord = bigPhoney.phonize(word)

            for phoneme in phonizedWord.lower().split(" "):
                isStressed = False
                if phoneme[-1].isdigit():
                    if phoneme[-1] == "1":
                        isStressed = True
                    phoneme = phoneme[:-1]

                try:
                    phoneme = phonemeMap[phoneme]
                except KeyError:
                    pass

                if phoneme[0] in vowels:
                    syllableCount += 1

                phonemes.append((phoneme, isStressed))

            for phoneme in phonemes:
                if capitalizeNext:
                    outputText += phoneme[0][0].upper()
                    capitalizeNext = False
                else:
                    outputText += phoneme[0][0]
                    if enableDiaeresis and (outputText[-2:] in digraphs or outputText[-3:] in trigraphs or (len(phoneme[0]) > 1 and outputText[-2:] + phoneme[0][1] in trigraphs)):
                        outputText += "\u0308"

                if enableStressMark and syllableCount > 1 and phoneme[1]:
                    outputText += "\u0301"

                outputText += phoneme[0][1:]

            word = ""

        if char in punctuation:
            capitalizeNext = True

        outputText += char

open("output.txt", "w").write(normalize("NFC", outputText))