Spaces:
Runtime error
Runtime error
Delete app.py
Browse files
app.py
DELETED
|
@@ -1,173 +0,0 @@
|
|
| 1 |
-
from flask import Flask, render_template, request, jsonify, Response
|
| 2 |
-
from werkzeug.utils import secure_filename
|
| 3 |
-
import os
|
| 4 |
-
from huggingface_hub import InferenceClient
|
| 5 |
-
import json
|
| 6 |
-
import base64
|
| 7 |
-
from io import BytesIO
|
| 8 |
-
import tempfile
|
| 9 |
-
from PIL import Image
|
| 10 |
-
import pytesseract
|
| 11 |
-
import fitz # PyMuPDF
|
| 12 |
-
import uuid
|
| 13 |
-
|
| 14 |
-
app = Flask(__name__)
|
| 15 |
-
app.secret_key = 'your-secret-key-here'
|
| 16 |
-
|
| 17 |
-
# Configuration
|
| 18 |
-
app.config['MAX_CONTENT_LENGTH'] = 16 * 1024 * 1024 # 16MB max file size
|
| 19 |
-
ALLOWED_EXTENSIONS = {'pdf'}
|
| 20 |
-
|
| 21 |
-
# Windows Configuration
|
| 22 |
-
#TESSERACT_PATH = r"D:\Python-files\tesseract\tesseract.exe" # Update this path
|
| 23 |
-
#pytesseract.pytesseract.tesseract_cmd = TESSERACT_PATH
|
| 24 |
-
|
| 25 |
-
# Store temporary file paths
|
| 26 |
-
temp_files = {}
|
| 27 |
-
|
| 28 |
-
def allowed_file(filename):
|
| 29 |
-
return '.' in filename and filename.rsplit('.', 1)[1].lower() in ALLOWED_EXTENSIONS
|
| 30 |
-
|
| 31 |
-
@app.route('/')
|
| 32 |
-
def index():
|
| 33 |
-
return render_template('index.html')
|
| 34 |
-
|
| 35 |
-
@app.route('/process_ai', methods=['POST'])
|
| 36 |
-
def process_ai():
|
| 37 |
-
try:
|
| 38 |
-
text = request.json.get('text', '')
|
| 39 |
-
image_data = request.json.get('image_data', None)
|
| 40 |
-
|
| 41 |
-
def generate():
|
| 42 |
-
client = InferenceClient()
|
| 43 |
-
|
| 44 |
-
if image_data:
|
| 45 |
-
# Process as image using chat completions
|
| 46 |
-
messages = [
|
| 47 |
-
{
|
| 48 |
-
"role": "user",
|
| 49 |
-
"content": [
|
| 50 |
-
{
|
| 51 |
-
"type": "text",
|
| 52 |
-
"text": "Analyze the MCQ question and provide the question and correct answer and if u want to give explanation just give in 2 to 3 lines:"
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"type": "image_url",
|
| 56 |
-
"image_url": {
|
| 57 |
-
"url": f"data:image/jpeg;base64,{image_data}"
|
| 58 |
-
}
|
| 59 |
-
}
|
| 60 |
-
]
|
| 61 |
-
}
|
| 62 |
-
]
|
| 63 |
-
|
| 64 |
-
stream = client.chat.completions.create(
|
| 65 |
-
model="meta-llama/Llama-3.2-11B-Vision-Instruct",
|
| 66 |
-
messages=messages,
|
| 67 |
-
max_tokens=500,
|
| 68 |
-
stream=True
|
| 69 |
-
)
|
| 70 |
-
else:
|
| 71 |
-
# Process as text using chat completions
|
| 72 |
-
messages = [
|
| 73 |
-
{
|
| 74 |
-
"role": "user",
|
| 75 |
-
"content": f"Analyze the MCQ question and provide the question and correct answer and if u want to give explanation just give in 2 to 3 lines:\n\n{text}"
|
| 76 |
-
}
|
| 77 |
-
]
|
| 78 |
-
|
| 79 |
-
stream = client.chat.completions.create(
|
| 80 |
-
model="Qwen/Qwen2.5-Coder-32B-Instruct",
|
| 81 |
-
messages=messages,
|
| 82 |
-
max_tokens=500,
|
| 83 |
-
stream=True
|
| 84 |
-
)
|
| 85 |
-
|
| 86 |
-
# Common streaming response handling for both text and image
|
| 87 |
-
for chunk in stream:
|
| 88 |
-
content = chunk.choices[0].delta.content
|
| 89 |
-
if content:
|
| 90 |
-
yield f'{{"chunk": {json.dumps(content)}}}\n'
|
| 91 |
-
|
| 92 |
-
return Response(generate(), mimetype='application/x-ndjson')
|
| 93 |
-
except Exception as e:
|
| 94 |
-
return jsonify({'error': str(e)}), 500
|
| 95 |
-
|
| 96 |
-
@app.route('/upload_pdf', methods=['POST', 'GET'])
|
| 97 |
-
def upload_pdf():
|
| 98 |
-
if request.method == 'POST':
|
| 99 |
-
if 'file' not in request.files:
|
| 100 |
-
return jsonify({'error': 'No file part'}), 400
|
| 101 |
-
|
| 102 |
-
file = request.files['file']
|
| 103 |
-
if file.filename == '' or not allowed_file(file.filename):
|
| 104 |
-
return jsonify({'error': 'Invalid file'}), 400
|
| 105 |
-
|
| 106 |
-
temp_pdf = None
|
| 107 |
-
try:
|
| 108 |
-
# Create a unique identifier for this upload
|
| 109 |
-
file_id = str(uuid.uuid4())
|
| 110 |
-
|
| 111 |
-
# Create a temporary file
|
| 112 |
-
temp_pdf = tempfile.NamedTemporaryFile(delete=False, suffix='.pdf')
|
| 113 |
-
file.save(temp_pdf.name)
|
| 114 |
-
temp_pdf.close()
|
| 115 |
-
|
| 116 |
-
# Store the temp file path with its ID
|
| 117 |
-
temp_files[file_id] = temp_pdf.name
|
| 118 |
-
|
| 119 |
-
return jsonify({'success': True, 'message': 'File uploaded', 'file_id': file_id})
|
| 120 |
-
|
| 121 |
-
except Exception as e:
|
| 122 |
-
if temp_pdf:
|
| 123 |
-
try:
|
| 124 |
-
os.unlink(temp_pdf.name)
|
| 125 |
-
except Exception as cleanup_error:
|
| 126 |
-
print(f"Warning: Could not delete temporary file during error cleanup: {cleanup_error}")
|
| 127 |
-
return jsonify({'error': str(e)}), 500
|
| 128 |
-
|
| 129 |
-
elif request.method == 'GET':
|
| 130 |
-
file_id = request.args.get('file_id')
|
| 131 |
-
if not file_id or file_id not in temp_files:
|
| 132 |
-
return jsonify({'error': 'No file to process'}), 400
|
| 133 |
-
|
| 134 |
-
temp_pdf_path = temp_files[file_id]
|
| 135 |
-
|
| 136 |
-
def generate():
|
| 137 |
-
try:
|
| 138 |
-
pdf_document = fitz.open(temp_pdf_path)
|
| 139 |
-
total_pages = pdf_document.page_count
|
| 140 |
-
texts = []
|
| 141 |
-
|
| 142 |
-
for page_num in range(total_pages):
|
| 143 |
-
page = pdf_document[page_num]
|
| 144 |
-
pix = page.get_pixmap()
|
| 145 |
-
img = Image.frombytes("RGB", [pix.width, pix.height], pix.samples)
|
| 146 |
-
text = pytesseract.image_to_string(img)
|
| 147 |
-
texts.append(text)
|
| 148 |
-
|
| 149 |
-
# Send progress update
|
| 150 |
-
progress = ((page_num + 1) / total_pages) * 100
|
| 151 |
-
yield f"data: {json.dumps({'progress': progress, 'page': page_num + 1, 'total': total_pages})}\n\n"
|
| 152 |
-
|
| 153 |
-
# Send final result
|
| 154 |
-
yield f"data: {json.dumps({'done': True, 'texts': texts})}\n\n"
|
| 155 |
-
|
| 156 |
-
# Clean up
|
| 157 |
-
pdf_document.close()
|
| 158 |
-
os.unlink(temp_pdf_path)
|
| 159 |
-
del temp_files[file_id]
|
| 160 |
-
|
| 161 |
-
except Exception as e:
|
| 162 |
-
print(f"Error during OCR: {e}")
|
| 163 |
-
if 'pdf_document' in locals():
|
| 164 |
-
pdf_document.close()
|
| 165 |
-
try:
|
| 166 |
-
os.unlink(temp_pdf_path)
|
| 167 |
-
del temp_files[file_id]
|
| 168 |
-
except:
|
| 169 |
-
pass
|
| 170 |
-
yield f"data: {json.dumps({'error': str(e)})}\n\n"
|
| 171 |
-
|
| 172 |
-
return Response(generate(), mimetype='text/event-stream')
|
| 173 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|