summarizeit / app.py
gnumanth's picture
Update app.py
3ebc257 verified
raw
history blame
4.4 kB
import gradio as gr
from markitdown import MarkItDown
import google.generativeai as genai
import tempfile
import os
from pathlib import Path
from transformers.utils import logging
# Initialize MarkItDown
md = MarkItDown()
# Configure Gemini AI
genai.configure(api_key=os.getenv('GEMINI_KEY'))
model = genai.GenerativeModel('gemini-2.0-flash-exp')
def process_with_markitdown(input_path):
"""Process file or URL with MarkItDown and return text content"""
try:
result = md.convert(input_path)
return result.text_content
except Exception as e:
return f"Error processing input: {str(e)}"
def save_uploaded_file(uploaded_file):
"""Saves an uploaded file to a temporary location."""
if uploaded_file is None:
return "No file uploaded."
try:
# Extract filename and file object from the tuple
filename, file_object = uploaded_file
temp_dir = tempfile.gettempdir()
temp_filename = os.path.join(temp_dir, filename)
with open(temp_filename, 'wb') as f:
f.write(file_object.read())
logger.info(filename, temp_filename)
return temp_filename
except Exception as e:
return f"An error occurred: {str(e)}"
async def summarize_text(text):
"""Summarize the input text using Gemini AI"""
try:
prompt = f"""Please provide a concise summary of the following text. Focus on the main points and key takeaways:
{text}
Summary:"""
response = await model.generate_content_async(prompt)
return response.text
except Exception as e:
return f"Error generating summary: {str(e)}"
async def process_input(input_text, uploaded_file=None):
"""Main function to process either URL or uploaded file"""
try:
if uploaded_file is not None:
# Handle file upload
temp_path = save_uploaded_file(uploaded_file)
if temp_path.startswith('Error'):
return temp_path
text = process_with_markitdown(temp_path)
# Clean up temporary file
try:
os.remove(temp_path)
except:
pass
elif input_text.startswith(('http://', 'https://')):
# Handle URL
text = process_with_markitdown(input_text)
else:
# Handle direct text input
text = input_text
if text.startswith('Error'):
return text
# Generate summary using Gemini AI
summary = await summarize_text(text)
return summary
except Exception as e:
return f"Error processing input: {str(e)}"
def clear_inputs():
return ["", None, ""]
# Create Gradio interface with drag-and-drop
with gr.Blocks(theme=gr.themes.Soft()) as iface:
gr.Markdown(
"""
# Summarizeit
> Summarize any document!
Enter a URL, paste text, or drag & drop a file to get a summary.
"""
)
with gr.Row():
input_text = gr.Textbox(
label="Enter URL or text",
placeholder="Enter a URL or paste text here...",
scale=2
)
with gr.Row():
file_upload = gr.File(
label="Drop files here or click to upload",
file_types=[
".pdf", ".docx", ".xlsx", ".csv", ".txt", ".md",
".html", ".htm", ".xml", ".json"
],
file_count="single",
scale=2
)
with gr.Row():
submit_btn = gr.Button("Summarize", variant="primary")
clear_btn = gr.Button("Clear")
output_text = gr.Textbox(
label="Summary",
lines=10,
show_copy_button=True
)
# Set up event handlers
submit_btn.click(
fn=process_input,
inputs=[input_text, file_upload],
outputs=output_text
)
clear_btn.click(
fn=clear_inputs,
outputs=[input_text, file_upload, output_text]
)
# Add examples
gr.Examples(
examples=[
["https://h3manth.com"],
["https://www.youtube.com/watch?v=bSHp7WVpPgc"],
["https://en.wikipedia.org/wiki/Three-body_problem"]
],
inputs=input_text
)
if __name__ == "__main__":
iface.launch()