app.py

import os
import subprocess
import sys
import gradio as gr
from transformers import pipeline
import spacy
import nltk
from nltk.corpus import wordnet

# Function to install GECToR
def install_gector():
    if not os.path.exists('gector'):
        print("Cloning GECToR repository...")
        subprocess.run(["git", "clone", "https://github.com/grammarly/gector.git"], check=True)
        
        # Install dependencies from GECToR requirements
        subprocess.run([sys.executable, "-m", "pip", "install", "-r", "gector/requirements.txt"], check=True)

        # Manually add GECToR to the Python path
        sys.path.append(os.path.abspath('gector'))

# Install and import GECToR
install_gector()

# Ensure the gector module path is added correctly
sys.path.insert(0, os.path.abspath('./gector'))

# Import GECToR after installation
from gector.gec_model import GecBERTModel

# Initialize GECToR model for grammar correction
gector_model = GecBERTModel(vocab_path='gector/data/output_vocabulary',
                            model_paths=['https://grammarly-nlp-data.s3.amazonaws.com/gector/roberta_1_gector.th'],
                            is_ensemble=False)

# Initialize the English text classification pipeline for AI detection
pipeline_en = pipeline(task="text-classification", model="Hello-SimpleAI/chatgpt-detector-roberta")

# Function to predict the label and score for English text (AI Detection)
def predict_en(text):
    res = pipeline_en(text)[0]
    return res['label'], res['score']

# Ensure necessary NLTK data is downloaded for Humanifier
nltk.download('wordnet')
nltk.download('omw-1.4')

# Ensure the SpaCy model is installed for Humanifier
try:
    nlp = spacy.load("en_core_web_sm")
except OSError:
    subprocess.run([sys.executable, "-m", "spacy", "download", "en_core_web_sm"])
    nlp = spacy.load("en_core_web_sm")

# Function to correct grammar using GECToR
def correct_grammar_with_gector(text):
    corrected_sentences = []
    sentences = [text]
    for sentence in sentences:
        preds = gector_model.handle_batch([sentence])
        corrected_sentences.append(preds[0])
    return ' '.join(corrected_sentences)

# Gradio app setup with three tabs
with gr.Blocks() as demo:
    with gr.Tab("AI Detection"):
        t1 = gr.Textbox(lines=5, label='Text')
        button1 = gr.Button("🤖 Predict!")
        label1 = gr.Textbox(lines=1, label='Predicted Label 🎃')
        score1 = gr.Textbox(lines=1, label='Prob')

        # Connect the prediction function to the button
        button1.click(predict_en, inputs=[t1], outputs=[label1, score1], api_name='predict_en')
    
    with gr.Tab("Humanifier"):
        text_input = gr.Textbox(lines=5, label="Input Text")
        paraphrase_button = gr.Button("Paraphrase & Correct")
        output_text = gr.Textbox(label="Paraphrased Text")

        # Connect the paraphrasing function to the button
        paraphrase_button.click(correct_grammar_with_gector, inputs=text_input, outputs=output_text)
    
    with gr.Tab("Grammar Correction"):
        grammar_input = gr.Textbox(lines=5, label="Input Text")
        grammar_button = gr.Button("Correct Grammar")
        grammar_output = gr.Textbox(label="Corrected Text")

        # Connect the GECToR grammar correction function to the button
        grammar_button.click(correct_grammar_with_gector, inputs=grammar_input, outputs=grammar_output)

# Launch the app with all functionalities
demo.launch()