from flask import Flask, render_template, request
from nltk.tokenize import word_tokenize
from nltk.corpus import wordnet
from nltk import pos_tag, ne_chunk
import textblob
from polyglot.detect import Detector
import numpy as np
from keras.models import load_model

app = Flask(__name__)
model = load_model("emotion_detector.h5")

@app.route("/")
def index():
    return render_template("index.html")

@app.route("/paraphrase", methods=["POST"])
def paraphrase():
    input_text = request.form["input_text"]
    options = request.form.getlist("options")

    # Remove special characters
    if "remove_special_characters" in options:
        input_text = remove_special_characters(input_text)
    
    # Correct grammar
    if "correct_grammar" in options:
        input_text = correct_grammar(input_text)

    # Summarize text
    if "summarize_text" in options:
        input_text = summarize_text(input_text)

    # Multilingual support
    target_language = request.form.get("target_language")
    if target_language:
        input_text = translate(input_text, target_language)

    # Custom synonyms
    custom_synonyms = request.form.getlist("custom_synonyms")
    for word, synonym in custom_synonyms:
        input_text = replace_word(input_text, word, synonym)

    # Output customization
    input_text = customise_output(input_text, options)

    # Integration with other NLP tools
    named_entities = get_named_entities(input_text)
    part_of_speech = get_part_of_speech(input_text)
    sentiment = get_sentiment(input_text)

    # Emotion detector
    emotion = detect_emotion(input_text)
    input_text = adjust_tone(input_text, emotion)

    return render_template("index.html", paraphrased_text=input_text, named_entities=named_entities, part_of_speech=part_of_speech, sentiment=sentiment)

def remove_special_characters(input_text):
    # Code to remove special characters
    return input_text

def summarize_text(input_text):
    # Code to summarize the text
    return input_text

def detect_language(input_text):
    detector = Detector(input_text)
    language = detector.language.code
    return language

def translate(input_text, target_language):
    blob = textblob.TextBlob(input_text)
    translated_text = blob.translate(to=target_language)
    return translated_text

def get_synonyms(word):
    synonyms = []
    for syn in wordnet.synsets(word):
        for lemma in syn.lemmas():
            synonyms.append(lemma.name())
    return synonyms

def replace_word(input_text, word, synonym):
    words = word_tokenize(input_text)
    words = [synonym if w == word else w for w in words]
    input_text = " ".join(words)
    return input_text

def customise_output(input_text, options):
    # Code to customise output based on options
    return input_text

def get_named_entities(input_text):
    named_entities = ne_chunk(pos_tag(word_tokenize(input_text)))
    return named_entities

def get_part_of_speech(input_text):
    pos = pos_tag(word_tokenize(input_text))
    return pos

def get_sentiment(input_text):
    blob = textblob.TextBlob(input_text)
    sentiment = blob.sentiment.polarity
    return sentiment

def correct_grammar(input_text):
    blob = textblob.TextBlob(input_text)
    corrected_text = str(blob.correct())
    return corrected_text

def detect_emotion(input_text):
    words = word_tokenize(input_text)
    words = [w.lower() for w in words]
    words = [w for w in words if w.isalpha()]
    input_text = " ".join(words)
    input_text = np.array([input_text])
    sentiment = model.predict(input_text, batch_size=1, verbose=0)[0]
    return sentiment

def adjust_tone(input_text, emotion):
    # Code to adjust tone based on emotion
    return input_text

if __name__ == "__main__":
    app.run(debug=True,port=7860,host="0.0.0.0")