Spaces:
Sleeping
Sleeping
File size: 16,652 Bytes
5753ed4 dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 5753ed4 dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 ed74fda dd5d745 5753ed4 dd5d745 ed74fda dd5d745 5753ed4 dd5d745 ed74fda dd5d745 ed74fda 5753ed4 ed74fda dd5d745 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 |
from flask import Flask, request, jsonify, send_from_directory
from werkzeug.utils import secure_filename
from werkzeug.security import generate_password_hash, check_password_hash
import pytesseract
from PIL import Image
import numpy as np
import faiss
import os
import pickle
from pdf2image import convert_from_bytes
import torch
import clip
import io
import json
import uuid
from datetime import datetime, timedelta
import jwt
import sqlite3
import tempfile
app = Flask(__name__)
app.config['SECRET_KEY'] = 'your-secret-key-change-this-in-production'
# Security configuration
SECRET_KEY = "your-secret-key-change-this-in-production"
ALGORITHM = "HS256"
ACCESS_TOKEN_EXPIRE_MINUTES = 30
# Set CLIP cache to writable directory
os.environ['CLIP_CACHE'] = '/app/clip_cache'
os.makedirs('/app/clip_cache', exist_ok=True)
# Directories
INDEX_PATH = "data/index.faiss"
LABELS_PATH = "data/labels.pkl"
DATABASE_PATH = "data/documents.db"
UPLOADS_DIR = "data/uploads"
os.makedirs("data", exist_ok=True)
os.makedirs("static", exist_ok=True)
os.makedirs(UPLOADS_DIR, exist_ok=True)
# Initialize database
def init_db():
conn = sqlite3.connect(DATABASE_PATH)
cursor = conn.cursor()
# Users table
cursor.execute('''
CREATE TABLE IF NOT EXISTS users (
id INTEGER PRIMARY KEY AUTOINCREMENT,
username TEXT UNIQUE NOT NULL,
password_hash TEXT NOT NULL,
is_active BOOLEAN DEFAULT TRUE
)
''')
# Documents table
cursor.execute('''
CREATE TABLE IF NOT EXISTS documents (
id TEXT PRIMARY KEY,
filename TEXT NOT NULL,
original_filename TEXT NOT NULL,
category TEXT NOT NULL,
similarity REAL NOT NULL,
ocr_text TEXT,
upload_date TEXT NOT NULL,
file_path TEXT NOT NULL
)
''')
# Insert default admin user if not exists
cursor.execute('SELECT * FROM users WHERE username = ?', ('admin',))
if not cursor.fetchone():
admin_hash = generate_password_hash('admin123')
cursor.execute('INSERT INTO users (username, password_hash) VALUES (?, ?)',
('admin', admin_hash))
conn.commit()
conn.close()
init_db()
# Initialize index and labels
index = faiss.IndexFlatL2(512)
labels = []
if os.path.exists(INDEX_PATH) and os.path.exists(LABELS_PATH):
try:
index = faiss.read_index(INDEX_PATH)
with open(LABELS_PATH, "rb") as f:
labels = pickle.load(f)
print(f"β
Loaded existing index with {len(labels)} labels")
except Exception as e:
print(f"β οΈ Failed to load existing index: {e}")
if os.path.exists(INDEX_PATH):
os.remove(INDEX_PATH)
if os.path.exists(LABELS_PATH):
os.remove(LABELS_PATH)
# Initialize CLIP model with custom cache
device = "cuda" if torch.cuda.is_available() else "cpu"
try:
clip_model, preprocess = clip.load("ViT-B/32", device=device, download_root='/app/clip_cache')
print("β
CLIP model loaded successfully")
except Exception as e:
print(f"β Failed to load CLIP model: {e}")
# Fallback initialization
clip_model = None
preprocess = None
# Helper functions
def save_index():
try:
faiss.write_index(index, INDEX_PATH)
with open(LABELS_PATH, "wb") as f:
pickle.dump(labels, f)
except Exception as e:
print(f"β Failed to save index: {e}")
def authenticate_user(username: str, password: str):
conn = sqlite3.connect(DATABASE_PATH)
cursor = conn.cursor()
cursor.execute('SELECT password_hash FROM users WHERE username = ? AND is_active = TRUE', (username,))
result = cursor.fetchone()
conn.close()
if result and check_password_hash(result[0], password):
return {"username": username}
return None
def create_access_token(data: dict):
expire = datetime.utcnow() + timedelta(minutes=ACCESS_TOKEN_EXPIRE_MINUTES)
to_encode = data.copy()
to_encode.update({"exp": expire})
return jwt.encode(to_encode, SECRET_KEY, algorithm=ALGORITHM)
def verify_token(token: str):
try:
payload = jwt.decode(token, SECRET_KEY, algorithms=[ALGORITHM])
username = payload.get("sub")
return username if username else None
except jwt.PyJWTError:
return None
def image_from_pdf(pdf_bytes):
try:
images = convert_from_bytes(pdf_bytes, dpi=200)
return images[0]
except Exception as e:
print(f"β PDF conversion error: {e}")
return None
def extract_text(image):
try:
if image.mode != 'RGB':
image = image.convert('RGB')
custom_config = r'--oem 3 --psm 6'
text = pytesseract.image_to_string(image, config=custom_config)
return text.strip() if text.strip() else "β No text detected"
except Exception as e:
return f"β OCR error: {str(e)}"
def get_clip_embedding(image):
try:
if clip_model is None:
return None
if image.mode != 'RGB':
image = image.convert('RGB')
image_input = preprocess(image).unsqueeze(0).to(device)
with torch.no_grad():
image_features = clip_model.encode_image(image_input)
image_features = image_features / image_features.norm(dim=-1, keepdim=True)
return image_features.cpu().numpy()[0]
except Exception as e:
print(f"β CLIP embedding error: {e}")
return None
def save_uploaded_file(file_content: bytes, filename: str) -> str:
file_id = str(uuid.uuid4())
file_extension = os.path.splitext(filename)[1]
saved_filename = f"{file_id}{file_extension}"
file_path = os.path.join(UPLOADS_DIR, saved_filename)
with open(file_path, 'wb') as f:
f.write(file_content)
return saved_filename
# Routes
@app.route("/")
def dashboard():
return send_from_directory('static', 'index.html')
@app.route("/static/<path:filename>")
def static_files(filename):
return send_from_directory('static', filename)
@app.route("/api/login", methods=["POST"])
def login():
username = request.form.get("username")
password = request.form.get("password")
user = authenticate_user(username, password)
if not user:
return jsonify({"detail": "Incorrect username or password"}), 401
access_token = create_access_token(data={"sub": user["username"]})
return jsonify({"access_token": access_token, "token_type": "bearer", "username": user["username"]})
@app.route("/api/upload-category", methods=["POST"])
def upload_category():
# Verify token
auth_header = request.headers.get('Authorization')
if not auth_header or not auth_header.startswith('Bearer '):
return jsonify({"error": "Missing or invalid token"}), 401
token = auth_header.split(' ')[1]
username = verify_token(token)
if not username:
return jsonify({"error": "Invalid token"}), 401
try:
label = request.form.get("label")
file = request.files.get("file")
if not label or not file:
return jsonify({"error": "Missing label or file"}), 400
file_content = file.read()
if file.content_type and file.content_type.startswith('application/pdf'):
image = image_from_pdf(file_content)
else:
image = Image.open(io.BytesIO(file_content))
if image is None:
return jsonify({"error": "Failed to process image"}), 400
embedding = get_clip_embedding(image)
if embedding is None:
return jsonify({"error": "Failed to generate embedding"}), 400
index.add(np.array([embedding]))
labels.append(label.strip())
save_index()
return jsonify({"message": f"β
Added category '{label}' (Total: {len(labels)} categories)", "status": "success"})
except Exception as e:
return jsonify({"error": str(e)}), 500
@app.route("/api/classify-document", methods=["POST"])
def classify_document():
# Verify token
auth_header = request.headers.get('Authorization')
if not auth_header or not auth_header.startswith('Bearer '):
return jsonify({"error": "Missing or invalid token"}), 401
token = auth_header.split(' ')[1]
username = verify_token(token)
if not username:
return jsonify({"error": "Invalid token"}), 401
try:
if len(labels) == 0:
return jsonify({"error": "No categories in database. Please add some first."}), 400
file = request.files.get("file")
if not file:
return jsonify({"error": "Missing file"}), 400
file_content = file.read()
if file.content_type and file.content_type.startswith('application/pdf'):
image = image_from_pdf(file_content)
else:
image = Image.open(io.BytesIO(file_content))
if image is None:
return jsonify({"error": "Failed to process image"}), 400
embedding = get_clip_embedding(image)
if embedding is None:
return jsonify({"error": "Failed to generate embedding"}), 400
k = min(3, len(labels))
D, I = index.search(np.array([embedding]), k=k)
if len(labels) > 0 and I[0][0] < len(labels):
similarity = 1 - D[0][0]
confidence_threshold = 0.35
best_match = labels[I[0][0]]
matches = []
for i in range(min(k, len(D[0]))):
if I[0][i] < len(labels):
sim = 1 - D[0][i]
matches.append({"category": labels[I[0][i]], "similarity": round(sim, 3)})
# Save classified document to SQLite
if similarity >= confidence_threshold:
saved_filename = save_uploaded_file(file_content, file.filename)
ocr_text = extract_text(image)
document_id = str(uuid.uuid4())
conn = sqlite3.connect(DATABASE_PATH)
cursor = conn.cursor()
cursor.execute('''
INSERT INTO documents (id, filename, original_filename, category, similarity, ocr_text, upload_date, file_path)
VALUES (?, ?, ?, ?, ?, ?, ?, ?)
''', (document_id, saved_filename, file.filename, best_match, round(similarity, 3),
ocr_text, datetime.now().isoformat(), os.path.join(UPLOADS_DIR, saved_filename)))
conn.commit()
conn.close()
return jsonify({
"status": "success",
"category": best_match,
"similarity": round(similarity, 3),
"confidence": "high",
"matches": matches,
"document_saved": True,
"document_id": document_id
})
else:
return jsonify({
"status": "low_confidence",
"category": best_match,
"similarity": round(similarity, 3),
"confidence": "low",
"matches": matches,
"document_saved": False
})
return jsonify({"error": "Document not recognized"}), 400
except Exception as e:
return jsonify({"error": str(e)}), 500
@app.route("/api/categories", methods=["GET"])
def get_categories():
# Verify token
auth_header = request.headers.get('Authorization')
if not auth_header or not auth_header.startswith('Bearer '):
return jsonify({"error": "Missing or invalid token"}), 401
token = auth_header.split(' ')[1]
username = verify_token(token)
if not username:
return jsonify({"error": "Invalid token"}), 401
categories = list(set(labels)) # Remove duplicates
category_counts = {}
for label in labels:
category_counts[label] = category_counts.get(label, 0) + 1
return jsonify({"categories": categories, "counts": category_counts})
@app.route("/api/documents/<category>", methods=["GET"])
def get_documents_by_category(category):
# Verify token
auth_header = request.headers.get('Authorization')
if not auth_header or not auth_header.startswith('Bearer '):
return jsonify({"error": "Missing or invalid token"}), 401
token = auth_header.split(' ')[1]
username = verify_token(token)
if not username:
return jsonify({"error": "Invalid token"}), 401
conn = sqlite3.connect(DATABASE_PATH)
cursor = conn.cursor()
cursor.execute('SELECT * FROM documents WHERE category = ? ORDER BY upload_date DESC', (category,))
documents = []
for row in cursor.fetchall():
documents.append({
"id": row[0],
"filename": row[1],
"original_filename": row[2],
"category": row[3],
"similarity": row[4],
"ocr_text": row[5],
"upload_date": row[6],
"file_path": row[7]
})
conn.close()
return jsonify({"documents": documents, "count": len(documents)})
@app.route("/api/documents/<document_id>", methods=["DELETE"])
def delete_document(document_id):
# Verify token
auth_header = request.headers.get('Authorization')
if not auth_header or not auth_header.startswith('Bearer '):
return jsonify({"error": "Missing or invalid token"}), 401
token = auth_header.split(' ')[1]
username = verify_token(token)
if not username:
return jsonify({"error": "Invalid token"}), 401
try:
conn = sqlite3.connect(DATABASE_PATH)
cursor = conn.cursor()
# Get document info first
cursor.execute('SELECT file_path FROM documents WHERE id = ?', (document_id,))
result = cursor.fetchone()
if not result:
conn.close()
return jsonify({"error": "Document not found"}), 404
file_path = result[0]
# Delete physical file
if file_path and os.path.exists(file_path):
os.remove(file_path)
# Delete from database
cursor.execute('DELETE FROM documents WHERE id = ?', (document_id,))
conn.commit()
conn.close()
return jsonify({"message": "Document deleted successfully", "status": "success"})
except Exception as e:
return jsonify({"error": str(e)}), 500
@app.route("/api/ocr", methods=["POST"])
def ocr_document():
# Verify token
auth_header = request.headers.get('Authorization')
if not auth_header or not auth_header.startswith('Bearer '):
return jsonify({"error": "Missing or invalid token"}), 401
token = auth_header.split(' ')[1]
username = verify_token(token)
if not username:
return jsonify({"error": "Invalid token"}), 401
try:
file = request.files.get("file")
if not file:
return jsonify({"error": "Missing file"}), 400
file_content = file.read()
if file.content_type and file.content_type.startswith('application/pdf'):
image = image_from_pdf(file_content)
else:
image = Image.open(io.BytesIO(file_content))
if image is None:
return jsonify({"error": "Failed to process image"}), 400
text = extract_text(image)
return jsonify({"text": text, "status": "success"})
except Exception as e:
return jsonify({"error": str(e)}), 500
@app.route("/api/stats", methods=["GET"])
def get_stats():
# Verify token
auth_header = request.headers.get('Authorization')
if not auth_header or not auth_header.startswith('Bearer '):
return jsonify({"error": "Missing or invalid token"}), 401
token = auth_header.split(' ')[1]
username = verify_token(token)
if not username:
return jsonify({"error": "Invalid token"}), 401
conn = sqlite3.connect(DATABASE_PATH)
cursor = conn.cursor()
cursor.execute('SELECT category, COUNT(*) FROM documents GROUP BY category')
category_stats = dict(cursor.fetchall())
cursor.execute('SELECT COUNT(*) FROM documents')
total_documents = cursor.fetchone()[0]
conn.close()
return jsonify({
"total_categories": len(set(labels)),
"total_documents": total_documents,
"category_distribution": category_stats
})
if __name__ == "__main__":
app.run(host="0.0.0.0", port=7860, debug=True)
|