# tools/readability_tool.py
import re
from textstat import textstat


def _count_syllables_german(word: str) -> int:
    """
    Sehr grobe Abschätzung der Silbenanzahl im Deutschen:
    Zähle Vokalgruppen (a, e, i, o, u, ä, ö, ü, y).
    """
    groups = re.findall(r'[aeiouyäöüy]+', word.lower())
    return max(1, len(groups))


def calculate_readability_score(text: str) -> dict:
    """
    Berechnet:
      - english_score: klassischer Flesch Reading Ease (Textstat-Library)
      - german_score: deutsche Flesch-Formel (ungefähr)
    Rückgabe: {"english_score": float, "german_score": float}
    """
    # English
    english_score = textstat.flesch_reading_ease(text)

    # German
    sentences = re.split(r'[.!?]+', text)
    sentences = [s for s in sentences if s.strip()]
    words = re.findall(r"\w+", text, flags=re.UNICODE)
    syllables = sum(_count_syllables_german(w) for w in words)

    if not sentences or not words:
        german_score = 0.0
    else:
        avg_sentence_length = len(words) / len(sentences)
        avg_syllables_per_word = syllables / len(words)
        # Deutsche Flesch-Formel (Annäherung):
        german_score = 180 - avg_sentence_length - (58.5 * avg_syllables_per_word)

    return {
        "english_score": round(english_score, 2),
        "german_score": round(german_score, 2)
    }