message

app.py

@@ -152,111 +152,116 @@ app = gr.mount_gradio_app(app, demo, path="/")
 def home():
     return RedirectResponse(url="/")
 """
-import gradio as gr
-import numpy as np
-import fitz  # PyMuPDF
-import torch
-from fastapi import FastAPI
+from fastapi import FastAPI, Form, File, UploadFile
+from fastapi.responses import RedirectResponse
+from fastapi.staticfiles import StaticFiles
+from pydantic import BaseModel
 from transformers import pipeline
+import os
 from PIL import Image
-from starlette.responses import RedirectResponse
-from openpyxl import load_workbook
-from docx import Document
-from pptx import Presentation
+import io
+import pdfplumber
+import docx
+import openpyxl
+import pytesseract
+from io import BytesIO
+import fitz  # PyMuPDF
+import easyocr
+from fastapi.templating import Jinja2Templates
+from starlette.requests import Request
 
-# ✅ Initialize FastAPI
+# Initialize the app
 app = FastAPI()
 
-# ✅ Check if CUDA is Available (For Debugging)
-device = "cpu"
-print(f"✅ Running on: {device}")
+# Mount the static directory to serve HTML, CSS, JS files
+app.mount("/static", StaticFiles(directory="static"), name="static")
 
-# ✅ Lazy Load Model Function (Loads Only When Needed)
-def get_qa_pipeline():
-    print("🔄 Loading QA Model on CPU...")
-    return pipeline("text-generation", model="TinyLlama/TinyLlama-1.1B-Chat-v1.0", device=-1)
+# Initialize transformers pipelines
+qa_pipeline = pipeline("question-answering", model="microsoft/phi-2", tokenizer="microsoft/phi-2")
+image_qa_pipeline = pipeline("image-question-answering", model="Salesforce/blip-vqa-base", tokenizer="Salesforce/blip-vqa-base")
 
-def get_image_captioning_pipeline():
-    print("🔄 Loading Image Captioning Model on CPU...")
-    return pipeline("image-to-text", model="nlpconnect/vit-gpt2-image-captioning", device=-1)
+# Initialize EasyOCR for image-based text extraction
+reader = easyocr.Reader(['en'])
 
-# ✅ File Type Validation
-ALLOWED_EXTENSIONS = {"pdf", "docx", "pptx", "xlsx"}
+# Define a template for rendering HTML
+templates = Jinja2Templates(directory="templates")
 
-def validate_file_type(file):
-    print(f"📂 Validating file: {file.name}")
-    ext = file.name.split(".")[-1].lower()
-    return None if ext in ALLOWED_EXTENSIONS else f"❌ Unsupported file format: {ext}"
-
-# ✅ Extract Text Functions (Optimized)
-def extract_text_from_pdf(file):
-    print("📄 Extracting text from PDF...")
-    with fitz.open(file.name) as doc:
-        return " ".join(page.get_text() for page in doc)
-
-def extract_text_from_docx(file):
-    print("📄 Extracting text from DOCX...")
-    doc = Document(file.name)
-    return " ".join(p.text for p in doc.paragraphs)
-
-def extract_text_from_pptx(file):
-    print("📄 Extracting text from PPTX...")
-    ppt = Presentation(file.name)
-    return " ".join(shape.text for slide in ppt.slides for shape in slide.shapes if hasattr(shape, "text"))
-
-def extract_text_from_excel(file):
-    print("📊 Extracting text from Excel...")
-    wb = load_workbook(file.name, data_only=True)
-    return " ".join(" ".join(str(cell) for cell in row if cell) for sheet in wb.worksheets for row in sheet.iter_rows(values_only=True))
-
-# ✅ Question Answering Function (Efficient Processing)
-async def answer_question(file, question: str):
-    print("🔍 Processing file for QA...")
-    
-    validation_error = validate_file_type(file)
-    if validation_error:
-        return validation_error
+# Function to process PDFs
+def extract_pdf_text(file_path: str):
+    with pdfplumber.open(file_path) as pdf:
+        text = ""
+        for page in pdf.pages:
+            text += page.extract_text()
+    return text
 
-    file_ext = file.name.split(".")[-1].lower()
+# Function to process DOCX files
+def extract_docx_text(file_path: str):
+    doc = docx.Document(file_path)
     text = ""
+    for para in doc.paragraphs:
+        text += para.text
+    return text
 
-    if file_ext == "pdf":
-        text = extract_text_from_pdf(file)
-    elif file_ext == "docx":
-        text = extract_text_from_docx(file)
-    elif file_ext == "pptx":
-        text = extract_text_from_pptx(file)
-    elif file_ext == "xlsx":
-        text = extract_text_from_excel(file)
-    
-    if not text.strip():
-        return "⚠️ No text extracted from the document."
-
-    print("✂️ Truncating text for faster processing...")
-    truncated_text = text[:1024]  # Reduce to 1024 characters for better speed
+# Function to process PPTX files
+def extract_pptx_text(file_path: str):
+    from pptx import Presentation
+    prs = Presentation(file_path)
+    text = ""
+    for slide in prs.slides:
+        for shape in slide.shapes:
+            if hasattr(shape, "text"):
+                text += shape.text
+    return text
 
-    qa_pipeline = get_qa_pipeline()
-    response = qa_pipeline(f"Question: {question}\nContext: {truncated_text}")
+# Function to extract text from images using OCR
+def extract_text_from_image(image: Image):
+    text = pytesseract.image_to_string(image)
+    return text
 
-    return response[0]["generated_text"]
+# Home route
+@app.get("/")
+def home():
+    return RedirectResponse(url="/docs")
+
+# Function to answer questions based on document content
+@app.post("/question-answering-doc")
+async def question_answering_doc(question: str = Form(...), file: UploadFile = File(...)):
+    # Save the uploaded file temporarily
+    file_path = f"temp_files/{file.filename}"
+    os.makedirs(os.path.dirname(file_path), exist_ok=True)
+    with open(file_path, "wb") as f:
+        f.write(await file.read())
+    
+    # Extract text based on file type
+    if file.filename.endswith(".pdf"):
+        text = extract_pdf_text(file_path)
+    elif file.filename.endswith(".docx"):
+        text = extract_docx_text(file_path)
+    elif file.filename.endswith(".pptx"):
+        text = extract_pptx_text(file_path)
+    else:
+        return {"error": "Unsupported file format"}
 
-# ✅ Gradio UI
-with gr.Blocks() as demo:
-    gr.Markdown("## 📄 AI-Powered Document & Image QA")
+    # Use the model for question answering
+    qa_result = qa_pipeline(question=question, context=text)
+    return {"answer": qa_result['answer']}
 
-    with gr.Row():
-        file_input = gr.File(label="Upload Document")
-        question_input = gr.Textbox(label="Ask a Question", placeholder="What is this document about?")
+# Function to answer questions based on images
+@app.post("/question-answering-image")
+async def question_answering_image(question: str = Form(...), image_file: UploadFile = File(...)):
+    # Open the uploaded image
+    image = Image.open(BytesIO(await image_file.read()))
     
-    answer_output = gr.Textbox(label="Answer")
-    submit_btn = gr.Button("Get Answer")
+    # Use EasyOCR to extract text if the image has textual content
+    image_text = extract_text_from_image(image)
     
-    submit_btn.click(answer_question, inputs=[file_input, question_input], outputs=answer_output)
-
-# ✅ Mount Gradio with FastAPI
-app = gr.mount_gradio_app(app, demo, path="/")
+    # Use the BLIP VQA model for question answering on the image
+    image_qa_result = image_qa_pipeline(image=image, question=question)
+    
+    return {"answer": image_qa_result['answer'], "image_text": image_text}
 
-@app.get("/")
-def home():
-    return RedirectResponse(url="/")
+# Serve the application in Hugging Face space
+@app.get("/docs")
+async def get_docs(request: Request):
+    return templates.TemplateResponse("static/index.html", {"request": request})
 

static/index.html

@@ -1,11 +1,32 @@
 <!DOCTYPE html>
-<html>
-    <head>
-    <link rel="stylesheet" href="/app.css">
-    </head>
-    <body>
+<html lang="en">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>AI Question Answering</title>
+</head>
+<body>
+    <h1>AI-powered Question Answering</h1>
+    <h3>Ask questions about documents or images</h3>
+
+    <form action="/question-answering-doc" method="POST" enctype="multipart/form-data">
+        <label for="question">Question:</label>
+        <input type="text" id="question" name="question" required><br><br>
         
+        <label for="file">Upload Document (PDF, DOCX, PPTX):</label>
+        <input type="file" id="file" name="file" required><br><br>
+
+        <input type="submit" value="Submit">
+    </form>
+
+    <form action="/question-answering-image" method="POST" enctype="multipart/form-data">
+        <label for="question">Question (Image-based):</label>
+        <input type="text" id="question" name="question" required><br><br>
 
-    </body>
+        <label for="image_file">Upload Image:</label>
+        <input type="file" id="image_file" name="image_file" accept="image/*" required><br><br>
 
-</html>
+        <input type="submit" value="Submit">
+    </form>
+</body>
+</html>