app.py

import gradio as gr
import os
import tempfile
from deep_translator import GoogleTranslator
from langdetect import detect
from fpdf import FPDF
import docx2txt

# Dictionary of language codes for translation
LANGUAGES = {
    "English": "en",
    "Spanish": "es",
    "French": "fr",
    "German": "de",
    "Italian": "it",
    "Portuguese": "pt",
    "Russian": "ru",
    "Chinese": "zh",
    "Japanese": "ja",
    "Korean": "ko",
    "Arabic": "ar",
    "Hindi": "hi"
}

def detect_language(text):
    try:
        return detect(text)
    except:
        return "en"  # Default to English if detection fails

def translate_text(text, target_language):
    if not text or not target_language:
        return "No text to translate"
        
    source_lang = detect_language(text)
    
    if source_lang == LANGUAGES[target_language]:
        return text
    
    try:
        translator = GoogleTranslator(source=source_lang, target=LANGUAGES[target_language])
        translation = translator.translate(text)
        return translation
    except Exception as e:
        return f"Translation failed: {str(e)}"

def extract_text_from_document(file):
    if not file:
        return ""
        
    file_path = file.name  # Gradio file object provides a 'name' attribute
    _, file_extension = os.path.splitext(file_path)
    
    if file_extension.lower() == '.txt':
        with open(file_path, 'r', encoding='utf-8', errors='replace') as f:
            return f.read()
    elif file_extension.lower() in ['.docx', '.doc']:
        return docx2txt.process(file_path)
    else:
        return "Unsupported file format. Please upload a .txt or .docx file."

def text_to_pdf(text, output_path):
    pdf = FPDF()
    pdf.add_page()
    pdf.set_font("Arial", size=12)
    
    lines = text.split('\n')
    for line in lines:
        try:
            pdf.multi_cell(0, 10, line.encode('latin-1', 'replace').decode('latin-1'))
        except Exception:
            pdf.multi_cell(0, 10, "[Text contains unsupported characters]")
    
    pdf.output(output_path)
    return output_path

def translate_and_save(input_text, input_file, target_language):
    if input_text:
        text_to_translate = input_text
    elif input_file is not None:
        text_to_translate = extract_text_from_document(input_file)
    else:
        return None, "Please provide either text or a document for translation."
    
    translated_text = translate_text(text_to_translate, target_language)
    
    with tempfile.NamedTemporaryFile(suffix=".pdf", delete=False) as temp_pdf:
        pdf_path = temp_pdf.name
    
    text_to_pdf(translated_text, pdf_path)
    
    return pdf_path, translated_text

# Create Gradio interface with explicit types
with gr.Blocks(title="Context-Aware Translation Tool") as demo:
    gr.Markdown("# Context-Aware Language Translation")
    gr.Markdown("This tool translates text while preserving context, idioms, and phrases.")
    
    with gr.Row():
        with gr.Column():
            input_text = gr.Textbox(label="Enter text to translate", lines=5, placeholder="Type your text here")
            input_file = gr.File(label="Or upload a document (.txt, .docx)", file_types=[".txt", ".docx"])
            target_language = gr.Dropdown(
                label="Target Language", 
                choices=list(LANGUAGES.keys()), 
                value="English"
            )
            translate_button = gr.Button("Translate")
        
        with gr.Column():
            output_text = gr.Textbox(label="Translation", lines=5, interactive=False)
            output_pdf = gr.File(label="Download as PDF")

    # Define the event without _js
    translate_button.click(
        fn=translate_and_save,
        inputs=[
            input_text,
            input_file,
            target_language
        ],
        outputs=[
            output_pdf,
            output_text
        ]
    )

if __name__ == "__main__":
    print(f"Gradio version: {gr.__version__}")
    demo.launch(debug=True)