Spaces:
Build error
Build error
Update app.py
Browse files
app.py
CHANGED
|
@@ -235,7 +235,184 @@ class EnhancedDocProcessor:
|
|
| 235 |
results.append({"status": "error", "error": str(e), "file": file_path})
|
| 236 |
return results
|
| 237 |
|
| 238 |
-
#
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 239 |
|
| 240 |
app = Flask(__name__)
|
| 241 |
processor = EnhancedDocProcessor()
|
|
|
|
| 235 |
results.append({"status": "error", "error": str(e), "file": file_path})
|
| 236 |
return results
|
| 237 |
|
| 238 |
+
# HTML template with embedded JavaScript
|
| 239 |
+
HTML_TEMPLATE = """
|
| 240 |
+
<!DOCTYPE html>
|
| 241 |
+
<html lang="en">
|
| 242 |
+
<head>
|
| 243 |
+
<meta charset="UTF-8">
|
| 244 |
+
<meta name="viewport" content="width=device-width, initial-scale=1.0">
|
| 245 |
+
<title>Document Processor</title>
|
| 246 |
+
<script src="https://cdn.tailwindcss.com"></script>
|
| 247 |
+
<style>
|
| 248 |
+
/* Additional custom styles can go here */
|
| 249 |
+
.processing {
|
| 250 |
+
animation: pulse 2s infinite;
|
| 251 |
+
}
|
| 252 |
+
@keyframes pulse {
|
| 253 |
+
0% { opacity: 1; }
|
| 254 |
+
50% { opacity: 0.5; }
|
| 255 |
+
100% { opacity: 1; }
|
| 256 |
+
}
|
| 257 |
+
</style>
|
| 258 |
+
</head>
|
| 259 |
+
<body class="bg-gray-50">
|
| 260 |
+
<div class="container mx-auto p-6 max-w-4xl">
|
| 261 |
+
<div class="mb-8">
|
| 262 |
+
<h1 class="text-3xl font-bold mb-2">Smart Document Processor</h1>
|
| 263 |
+
<p class="text-gray-600">Upload and analyze PDF documents with AI</p>
|
| 264 |
+
</div>
|
| 265 |
+
|
| 266 |
+
<!-- Upload Section -->
|
| 267 |
+
<div class="mb-8">
|
| 268 |
+
<div id="dropZone" class="border-2 border-dashed border-gray-300 rounded-lg p-8 text-center hover:border-blue-500 transition-colors">
|
| 269 |
+
<input type="file" multiple accept=".pdf" id="fileInput" class="hidden">
|
| 270 |
+
<div class="cursor-pointer">
|
| 271 |
+
<svg class="w-12 h-12 text-gray-400 mx-auto mb-4" xmlns="http://www.w3.org/2000/svg" fill="none" viewBox="0 0 24 24" stroke="currentColor">
|
| 272 |
+
<path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M7 16a4 4 0 01-.88-7.903A5 5 0 1115.9 6L16 6a5 5 0 011 9.9M15 13l-3-3m0 0l-3 3m3-3v12"/>
|
| 273 |
+
</svg>
|
| 274 |
+
<span class="text-lg mb-2 block">Drop PDFs here or click to upload</span>
|
| 275 |
+
<span class="text-sm text-gray-500">Supports multiple files</span>
|
| 276 |
+
</div>
|
| 277 |
+
</div>
|
| 278 |
+
</div>
|
| 279 |
+
|
| 280 |
+
<!-- File List -->
|
| 281 |
+
<div id="fileList" class="mb-8 hidden">
|
| 282 |
+
<h2 class="text-xl font-semibold mb-4">Selected Files</h2>
|
| 283 |
+
<div id="fileListContent" class="space-y-2"></div>
|
| 284 |
+
<button id="processButton" class="mt-4 bg-blue-600 text-white px-6 py-2 rounded-lg hover:bg-blue-700 disabled:opacity-50">
|
| 285 |
+
Process Documents
|
| 286 |
+
</button>
|
| 287 |
+
</div>
|
| 288 |
+
|
| 289 |
+
<!-- Results Section -->
|
| 290 |
+
<div id="results" class="space-y-4"></div>
|
| 291 |
+
|
| 292 |
+
<!-- Error Alert -->
|
| 293 |
+
<div id="error" class="hidden mt-4 bg-red-100 border border-red-400 text-red-700 px-4 py-3 rounded"></div>
|
| 294 |
+
</div>
|
| 295 |
+
|
| 296 |
+
<script>
|
| 297 |
+
let files = [];
|
| 298 |
+
const dropZone = document.getElementById('dropZone');
|
| 299 |
+
const fileInput = document.getElementById('fileInput');
|
| 300 |
+
const fileList = document.getElementById('fileList');
|
| 301 |
+
const fileListContent = document.getElementById('fileListContent');
|
| 302 |
+
const processButton = document.getElementById('processButton');
|
| 303 |
+
const resultsDiv = document.getElementById('results');
|
| 304 |
+
const errorDiv = document.getElementById('error');
|
| 305 |
+
|
| 306 |
+
// Drag and drop handlers
|
| 307 |
+
dropZone.addEventListener('dragover', (e) => {
|
| 308 |
+
e.preventDefault();
|
| 309 |
+
dropZone.classList.add('border-blue-500');
|
| 310 |
+
});
|
| 311 |
+
|
| 312 |
+
dropZone.addEventListener('dragleave', () => {
|
| 313 |
+
dropZone.classList.remove('border-blue-500');
|
| 314 |
+
});
|
| 315 |
+
|
| 316 |
+
dropZone.addEventListener('drop', (e) => {
|
| 317 |
+
e.preventDefault();
|
| 318 |
+
dropZone.classList.remove('border-blue-500');
|
| 319 |
+
handleFiles(e.dataTransfer.files);
|
| 320 |
+
});
|
| 321 |
+
|
| 322 |
+
dropZone.addEventListener('click', () => {
|
| 323 |
+
fileInput.click();
|
| 324 |
+
});
|
| 325 |
+
|
| 326 |
+
fileInput.addEventListener('change', (e) => {
|
| 327 |
+
handleFiles(e.target.files);
|
| 328 |
+
});
|
| 329 |
+
|
| 330 |
+
function handleFiles(uploadedFiles) {
|
| 331 |
+
files = Array.from(uploadedFiles).filter(file => file.name.toLowerCase().endsWith('.pdf'));
|
| 332 |
+
updateFileList();
|
| 333 |
+
}
|
| 334 |
+
|
| 335 |
+
function updateFileList() {
|
| 336 |
+
if (files.length > 0) {
|
| 337 |
+
fileList.classList.remove('hidden');
|
| 338 |
+
fileListContent.innerHTML = files.map((file, index) => `
|
| 339 |
+
<div class="flex items-center p-3 bg-gray-50 rounded">
|
| 340 |
+
<svg class="w-5 h-5 text-gray-500 mr-3" xmlns="http://www.w3.org/2000/svg" fill="none" viewBox="0 0 24 24" stroke="currentColor">
|
| 341 |
+
<path stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M9 12h6m-6 4h6m2 5H7a2 2 0 01-2-2V5a2 2 0 012-2h5.586a1 1 0 01.707.293l5.414 5.414a1 1 0 01.293.707V19a2 2 0 01-2 2z"/>
|
| 342 |
+
</svg>
|
| 343 |
+
<span>${file.name}</span>
|
| 344 |
+
</div>
|
| 345 |
+
`).join('');
|
| 346 |
+
} else {
|
| 347 |
+
fileList.classList.add('hidden');
|
| 348 |
+
}
|
| 349 |
+
}
|
| 350 |
+
|
| 351 |
+
|
| 352 |
+
processButton.addEventListener('click', async () => {
|
| 353 |
+
if (files.length === 0) return;
|
| 354 |
+
|
| 355 |
+
processButton.disabled = true;
|
| 356 |
+
processButton.innerHTML = `
|
| 357 |
+
<svg class="animate-spin -ml-1 mr-3 h-5 w-5 text-white inline" xmlns="http://www.w3.org/2000/svg" fill="none" viewBox="0 0 24 24">
|
| 358 |
+
<circle class="opacity-25" cx="12" cy="12" r="10" stroke="currentColor" stroke-width="4"></circle>
|
| 359 |
+
<path class="opacity-75" fill="currentColor" d="M4 12a8 8 0 018-8V0C5.373 0 0 5.373 0 12h4zm2 5.291A7.962 7.962 0 014 12H0c0 3.042 1.135 5.824 3 7.938l3-2.647z"></path>
|
| 360 |
+
</svg>
|
| 361 |
+
Processing...
|
| 362 |
+
`;
|
| 363 |
+
|
| 364 |
+
const formData = new FormData();
|
| 365 |
+
files.forEach(file => {
|
| 366 |
+
formData.append('files[]', file);
|
| 367 |
+
});
|
| 368 |
+
|
| 369 |
+
try {
|
| 370 |
+
const response = await fetch('/batch_process', {
|
| 371 |
+
method: 'POST',
|
| 372 |
+
body: formData
|
| 373 |
+
});
|
| 374 |
+
|
| 375 |
+
const data = await response.json();
|
| 376 |
+
displayResults(data);
|
| 377 |
+
errorDiv.classList.add('hidden');
|
| 378 |
+
} catch (error) {
|
| 379 |
+
errorDiv.textContent = 'Failed to process documents. Please try again.';
|
| 380 |
+
errorDiv.classList.remove('hidden');
|
| 381 |
+
} finally {
|
| 382 |
+
processButton.disabled = false;
|
| 383 |
+
processButton.textContent = 'Process Documents';
|
| 384 |
+
}
|
| 385 |
+
});
|
| 386 |
+
|
| 387 |
+
function displayResults(results) {
|
| 388 |
+
resultsDiv.innerHTML = results.map(result => `
|
| 389 |
+
<div class="border rounded-lg p-4 bg-white shadow-sm">
|
| 390 |
+
<h3 class="font-medium mb-2">${result.result.filename}</h3>
|
| 391 |
+
<div class="grid grid-cols-2 gap-4">
|
| 392 |
+
<div>
|
| 393 |
+
<span class="text-gray-600">Type:</span>
|
| 394 |
+
<span class="ml-2">${result.result.doc_type}</span>
|
| 395 |
+
</div>
|
| 396 |
+
<div>
|
| 397 |
+
<span class="text-gray-600">Date:</span>
|
| 398 |
+
<span class="ml-2">${result.result.date || 'N/A'}</span>
|
| 399 |
+
</div>
|
| 400 |
+
<div>
|
| 401 |
+
<span class="text-gray-600">Amount:</span>
|
| 402 |
+
<span class="ml-2">${result.result.amount ? '$' + result.result.amount.toFixed(2) : 'N/A'}</span>
|
| 403 |
+
</div>
|
| 404 |
+
<div>
|
| 405 |
+
<span class="text-gray-600">Person:</span>
|
| 406 |
+
<span class="ml-2">${result.result.person_name}</span>
|
| 407 |
+
</div>
|
| 408 |
+
</div>
|
| 409 |
+
</div>
|
| 410 |
+
`).join('');
|
| 411 |
+
}
|
| 412 |
+
</script>
|
| 413 |
+
</body>
|
| 414 |
+
</html>
|
| 415 |
+
"""
|
| 416 |
|
| 417 |
app = Flask(__name__)
|
| 418 |
processor = EnhancedDocProcessor()
|