import streamlit as st
import nltk
from nltk.corpus import cmudict
from difflib import SequenceMatcher

# Load CMU Pronouncing Dictionary
nltk.download('cmudict') 
d = cmudict.dict()

# Function to get phonetic transcription of a word
def phonetic_transcription(word):
    try:
        return d[word.lower()][0]
    except KeyError:
        return None

# Function to calculate phonetic similarity between two words
def phonetic_similarity(word1, word2):
    pt1 = phonetic_transcription(word1)
    pt2 = phonetic_transcription(word2)
    if pt1 is None or pt2 is None:
        return 0
    else:
        return SequenceMatcher(None, pt1, pt2).ratio()

# User input for list of words and similarity threshold
words = st.text_input("Enter list of words (separated by commas):")
threshold = st.slider("Similarity threshold:", min_value=0.0, max_value=1.0, value=0.5)

if words:
    words = [word.strip() for word in words.split(",")]
    n_words = len(words)

    # Calculate phonetic similarity matrix
    similarity_matrix = [[0 for _ in range(n_words)] for _ in range(n_words)]
    for i in range(n_words):
        for j in range(i+1, n_words):
            similarity = phonetic_similarity(words[i], words[j])
            similarity_matrix[i][j] = similarity
            similarity_matrix[j][i] = similarity

    # Find similar words based on similarity threshold
    similar_words = []
    for i in range(n_words):
        similar_words.append([words[j] for j in range(n_words) if similarity_matrix[i][j] >= threshold])

    # Display similar words with matching score
    for i in range(n_words):
        st.write(f"{words[i]}: {[f'{word} ({int(similarity_matrix[i][j]*100)}%)' for j, word in enumerate(similar_words[i])]}")