"""
Implementation of phonem
Based on:
Wilde, Georg ; Meyer, Carsten: Doppelgänger gesucht - Ein Programm fur
kontext-sensitive phonetische Textumwandlung. In: ct Magazin fur
Computer & Technik 25 (1988)
Inspired by Talisman's implementation:
https://raw.githubusercontent.com/Yomguithereal/talisman/4c8ac7fd9/src/phonetics/german/phonem.js
"""
import itertools
from typing import Dict, Union
SUBSTITUTIONS = {
"SC": "C",
"SZ": "C",
"CZ": "C",
"TZ": "C",
"TS": "C",
"KS": "X",
"PF": "V",
"PH": "V",
"QU": "KW",
"UE": "Y",
"AE": "E",
"OE": "Ö",
"EI": "AY",
"EY": "AY",
"EU": "OY",
"AU": "A§",
"OU": "§",
}
TRANSLATIONS: Dict[str, Union[int, str, None]] = {
"Z": "C",
"K": "C",
"G": "C",
"Q": "C",
"Ç": "C",
"Ñ": "N",
"ß": "S",
"F": "V",
"W": "V",
"P": "B",
"T": "D",
"Á": "A",
"À": "A",
"Â": "A",
"Ã": "A",
"Å": "A",
"Ä": "E",
"Æ": "E",
"É": "E",
"È": "E",
"Ê": "E",
"Ë": "E",
"I": "Y",
"J": "Y",
"Ì": "Y",
"Í": "Y",
"Î": "Y",
"Ï": "Y",
"Ü": "Y",
"Ý": "Y",
"§": "U",
"Ú": "U",
"Ù": "U",
"Û": "U",
"Ô": "O",
"Ò": "O",
"Ó": "O",
"Õ": "O",
"Ø": "Ö",
}
ACCEPTABLE_LETTERS = "ABCDLMNORSUVWXYÖ"
WHITELIST_LETTERS = set("ABCDEFGHIJKLMNOPQRSTUVWXYZ -")
WHITELIST_LETTERS.update(TRANSLATIONS.keys() | SUBSTITUTIONS.keys())
WHITELIST_LETTERS.discard(
"§"
) # Do not allow this non-printable character in original name
[docs]
def phonem_encode(string: str) -> str:
"""
Implements the phonem phonetic algorithm suitable for Germanic names.
Based on:
Wilde, Georg ; Meyer, Carsten: Doppelgänger gesucht - Ein Programm fur
kontext-sensitive phonetische Textumwandlung. In: ct Magazin fur
Computer & Technik 25 (1988)
Inspired by Talisman's implementation:
https://raw.githubusercontent.com/Yomguithereal/talisman/4c8ac7fd9/src/phonetics/german/phonem.js
Spaces and dashes (' ', '-'), are considered as word/name separators.
:param string: string to encode
:return: encoded string
"""
# Strip all but acceptable letters and separators
string = "".join(filter(lambda _: _ in WHITELIST_LETTERS, string.upper()))
# Substitute
for match, replacement in SUBSTITUTIONS.items():
string = string.replace(match, replacement)
# Translate
string = string.translate(str.maketrans(TRANSLATIONS))
# Squeeze and filter unacceptable letters
output = "".join(
char for char, _ in itertools.groupby(string) if char in ACCEPTABLE_LETTERS
)
return output