INNER CODE UNIT · Python

normalize_language_label

dedsec1121fk/DedSec · Scripts/DedSec Market.py:396

def normalize_language_label(text):
    value = strip_markdown(str(text or ""))
    value = html_lib.unescape(value).replace("\xa0", " ").strip().lower()
    value = remove_accents(value)
    value = re.sub(r"\s+", " ", value)
    return value.strip(" \t\r\n#>*_`[](){}.!?")


def detect_language_label(text):
    value = normalize_language_label(text)
    if not value:
        return None

    english_aliases = {"english", "en", "eng", "αγγλικα", "αγγλικη εκδοση", "🇬🇧", "🇺🇸"}
    greek_aliases = {"greek", "el", "gr", "ellinika", "ελληνικα", "ελληνικη εκδοση", "🇬🇷"}

    if value in english_aliases:
        return "en"

View source record →