Spaces:
Running
Running
File size: 24,421 Bytes
5e1a30c 1cdeab3 5e1a30c 1cdeab3 5e1a30c |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 |
"""
Database Manager for Epic 2 Demo Persistent Storage
==================================================
Handles database connections, operations, and high-level persistence management
for the Epic 2 demo to achieve <5 second initialization times.
"""
import logging
import hashlib
import time
import uuid
from datetime import datetime, timedelta
from pathlib import Path
from typing import Dict, Any, List, Optional, Tuple, Union
from contextlib import contextmanager
import numpy as np
from sqlalchemy import create_engine, text
from sqlalchemy.orm import sessionmaker, Session
from sqlalchemy.exc import SQLAlchemyError
from sqlalchemy.pool import StaticPool
from .database_schema import Base, Document, DocumentChunk, SystemCache, ProcessingSession, DatabaseSchema
logger = logging.getLogger(__name__)
class DatabaseManager:
"""Manages database operations for Epic 2 demo persistence"""
def __init__(self, database_url: str = "sqlite:///demo/epic2_demo.db", echo: bool = False):
"""
Initialize database manager
Args:
database_url: Database connection URL
echo: Whether to echo SQL statements (for debugging)
"""
self.database_url = database_url
self.echo = echo
# Create database directory if using SQLite
if database_url.startswith("sqlite:///"):
db_path = Path(database_url.replace("sqlite:///", ""))
db_path.parent.mkdir(parents=True, exist_ok=True)
# Create engine with optimized settings
self.engine = create_engine(
database_url,
echo=echo,
poolclass=StaticPool if "sqlite" in database_url else None,
connect_args={"check_same_thread": False} if "sqlite" in database_url else {},
pool_pre_ping=True,
pool_recycle=3600 # 1 hour
)
# Create session factory
self.SessionLocal = sessionmaker(
autocommit=False,
autoflush=False,
bind=self.engine
)
# Initialize database
self._initialize_database()
def _initialize_database(self) -> None:
"""Initialize database tables and indexes"""
try:
logger.info("Initializing database schema...")
DatabaseSchema.create_all_tables(self.engine)
# Optimize SQLite if using it
if "sqlite" in self.database_url:
self._optimize_sqlite()
logger.info("Database initialization complete")
except Exception as e:
logger.error(f"Database initialization failed: {e}")
raise
def _optimize_sqlite(self) -> None:
"""Apply SQLite-specific optimizations"""
try:
with self.engine.connect() as conn:
# Performance optimizations
conn.execute(text("PRAGMA journal_mode = WAL"))
conn.execute(text("PRAGMA synchronous = NORMAL"))
conn.execute(text("PRAGMA cache_size = 10000"))
conn.execute(text("PRAGMA temp_store = MEMORY"))
conn.execute(text("PRAGMA mmap_size = 268435456")) # 256MB
conn.commit()
logger.info("SQLite optimizations applied")
except Exception as e:
logger.warning(f"SQLite optimization failed: {e}")
@contextmanager
def get_session(self):
"""Context manager for database sessions"""
session = self.SessionLocal()
try:
yield session
session.commit()
except Exception as e:
session.rollback()
logger.error(f"Database session error: {e}")
raise
finally:
session.close()
def get_database_stats(self) -> Dict[str, Any]:
"""Get comprehensive database statistics"""
with self.get_session() as session:
stats = DatabaseSchema.get_database_stats(session)
# Add database file size if SQLite
if "sqlite" in self.database_url:
try:
db_path = Path(self.database_url.replace("sqlite:///", ""))
if db_path.exists():
stats['database_size_mb'] = db_path.stat().st_size / (1024 * 1024)
except:
pass
return stats
def is_database_populated(self) -> bool:
"""Check if database has any processed documents"""
try:
with self.get_session() as session:
count = session.query(Document).filter(
Document.processing_status == 'completed'
).count()
return count > 0
except:
return False
def is_cache_valid(self, pdf_files: List[Path], processor_config: Dict[str, Any],
embedder_config: Dict[str, Any]) -> bool:
"""
Check if database cache is valid for given files and configurations
Args:
pdf_files: List of PDF files to check
processor_config: Document processor configuration
embedder_config: Embedder configuration
Returns:
True if cache is valid and up-to-date
"""
try:
with self.get_session() as session:
# Simple check: do we have any completed documents in database?
total_docs = session.query(Document).filter(
Document.processing_status == 'completed'
).count()
total_chunks = session.query(DocumentChunk).filter(
DocumentChunk.embedding_vector != None
).count()
logger.info(f"Database validation: {total_docs} documents, {total_chunks} chunks with embeddings")
if total_docs == 0 or total_chunks == 0:
logger.info("No valid documents/chunks in database")
return False
# Check if we have any matching files
available_files = session.query(Document.filename).filter(
Document.processing_status == 'completed'
).all()
available_filenames = [doc.filename for doc in available_files]
requested_filenames = [pdf_file.name for pdf_file in pdf_files]
# Option A: If no corpus files requested but database has documents, load ALL
if len(requested_filenames) == 0 and len(available_filenames) > 0:
logger.info(f"No corpus files specified, but database has {len(available_filenames)} documents - will load ALL")
return True
matching_files = [f for f in requested_filenames if f in available_filenames]
logger.info(f"File matching: {len(matching_files)}/{len(requested_filenames)} files available in database")
# Accept if we have at least some matching files
if len(matching_files) > 0:
logger.info("Database cache validation successful (partial match)")
return True
else:
logger.info("No matching files in database")
return False
except Exception as e:
logger.error(f"Cache validation error: {e}")
return False
def load_documents_and_embeddings(self, pdf_files: List[Path]) -> Tuple[List[Any], Optional[np.ndarray]]:
"""
Load documents and embeddings from database
Args:
pdf_files: List of PDF files to load
Returns:
Tuple of (documents, embeddings) or (None, None) if failed
"""
try:
with self.get_session() as session:
# Load all chunks for the specified files
file_names = [f.name for f in pdf_files]
# First check if we have any documents at all
total_docs = session.query(Document).count()
logger.info(f"Total documents in database: {total_docs}")
if total_docs == 0:
logger.warning("No documents found in database")
return None, None
# Check which files we have
available_docs = session.query(Document.filename).filter(
Document.processing_status == 'completed'
).all()
available_files = [doc.filename for doc in available_docs]
logger.info(f"Available files in database: {available_files[:5]}...") # Show first 5
# Option A: If no specific files requested, load ALL documents
if len(file_names) == 0:
logger.info(f"No specific files requested - loading ALL {len(available_files)} documents from database")
chunks = session.query(DocumentChunk).join(Document).filter(
Document.processing_status == 'completed',
DocumentChunk.embedding_vector != None
).order_by(Document.id, DocumentChunk.chunk_index).all()
else:
# Find intersection of requested and available files
matching_files = [f for f in file_names if f in available_files]
logger.info(f"Matching files: {len(matching_files)}/{len(file_names)}")
if not matching_files:
logger.warning("No matching files found in database")
return None, None
chunks = session.query(DocumentChunk).join(Document).filter(
Document.filename.in_(matching_files),
Document.processing_status == 'completed',
DocumentChunk.embedding_vector != None
).order_by(Document.id, DocumentChunk.chunk_index).all()
if not chunks:
logger.warning("No chunks found in database")
return None, None
# Convert chunks to document objects and collect embeddings
documents = []
embeddings = []
for chunk in chunks:
# Create document-like object
doc = {
'id': chunk.id,
'content': chunk.content,
'metadata': chunk.chunk_metadata or {},
'confidence': chunk.confidence_score or 0.8,
'embedding': chunk.get_embedding()
}
# Add document metadata
if doc['metadata'] is None:
doc['metadata'] = {}
doc['metadata'].update({
'source': chunk.document.filename,
'page': chunk.chunk_metadata.get('page', 1) if chunk.chunk_metadata else 1,
'chunk_index': chunk.chunk_index
})
documents.append(doc)
# Collect embedding
embedding = chunk.get_embedding()
if embedding is not None:
embeddings.append(embedding)
else:
logger.warning(f"Missing embedding for chunk {chunk.id}")
if not embeddings:
logger.warning("No embeddings found in database")
return documents, None
embeddings_array = np.array(embeddings)
logger.info(f"Loaded {len(documents)} documents and {embeddings_array.shape} embeddings from database")
return documents, embeddings_array
except Exception as e:
logger.error(f"Failed to load from database: {e}")
return None, None
def save_documents_and_embeddings(self, documents: List[Any], pdf_files: List[Path],
processor_config: Dict[str, Any], embedder_config: Dict[str, Any]) -> bool:
"""
Save documents and embeddings to database
Args:
documents: List of processed document objects
pdf_files: List of source PDF files
processor_config: Document processor configuration
embedder_config: Embedder configuration
Returns:
True if save successful
"""
try:
processor_hash = self._hash_config(processor_config)
embedder_hash = self._hash_config(embedder_config)
# Create processing session
session_id = str(uuid.uuid4())
processing_start = time.time()
with self.get_session() as session:
# Create processing session record
proc_session = ProcessingSession(
session_id=session_id,
processor_config_hash=processor_hash,
embedder_config_hash=embedder_hash,
documents_processed=len(pdf_files),
chunks_created=len(documents)
)
session.add(proc_session)
session.flush()
# Group documents by source file
docs_by_file = {}
for doc in documents:
# Get source and extract filename
metadata = doc.get('metadata', {}) if isinstance(doc, dict) else getattr(doc, 'metadata', {})
source = metadata.get('source', 'unknown')
# Extract filename from full path
import os
if source != 'unknown':
source_filename = os.path.basename(source)
else:
source_filename = metadata.get('source_name', 'unknown')
if source_filename not in docs_by_file:
docs_by_file[source_filename] = []
docs_by_file[source_filename].append(doc)
logger.info(f"Grouped documents by file: {list(docs_by_file.keys())[:5]}...") # Show first 5
# Process each file
for pdf_file in pdf_files:
file_docs = docs_by_file.get(pdf_file.name, [])
if not file_docs:
logger.warning(f"No documents found for file: {pdf_file.name}")
continue
# Create or update document record
file_hash = self._hash_file(pdf_file)
file_mtime = pdf_file.stat().st_mtime
doc_record = session.query(Document).filter(
Document.filename == pdf_file.name
).first()
if not doc_record:
doc_record = Document(
filename=pdf_file.name,
file_path=str(pdf_file),
file_hash=file_hash,
file_size=pdf_file.stat().st_size,
file_mtime=file_mtime,
processor_config_hash=processor_hash,
chunk_count=len(file_docs),
processing_status='completed',
doc_metadata={} # Initialize with empty metadata
)
session.add(doc_record)
session.flush()
else:
# Update existing record
doc_record.file_hash = file_hash
doc_record.file_mtime = file_mtime
doc_record.processor_config_hash = processor_hash
doc_record.chunk_count = len(file_docs)
doc_record.processing_status = 'completed'
doc_record.processed_at = datetime.utcnow()
# Delete old chunks
session.query(DocumentChunk).filter(
DocumentChunk.document_id == doc_record.id
).delete()
# Save chunks
for idx, doc in enumerate(file_docs):
# Get content and metadata properly
if isinstance(doc, dict):
content = doc.get('content', '')
metadata = doc.get('metadata', {})
confidence = doc.get('confidence', 0.8)
else:
content = getattr(doc, 'content', '')
metadata = getattr(doc, 'metadata', {})
confidence = getattr(doc, 'confidence', 0.8)
chunk = DocumentChunk(
document_id=doc_record.id,
chunk_index=idx,
content=content,
content_hash=self._hash_text(content),
chunk_metadata=metadata,
embedding_model=embedder_config.get('model', {}).get('model_name', 'unknown'),
embedder_config_hash=embedder_hash,
confidence_score=confidence
)
# Set embedding if available
embedding = None
if hasattr(doc, 'embedding') and doc.embedding is not None:
embedding = doc.embedding
elif isinstance(doc, dict) and 'embedding' in doc and doc['embedding'] is not None:
embedding = doc['embedding']
if embedding is not None:
# Convert to numpy array if it's a list
if isinstance(embedding, list):
embedding = np.array(embedding, dtype=np.float32)
elif not isinstance(embedding, np.ndarray):
embedding = np.array(embedding, dtype=np.float32)
chunk.set_embedding(embedding)
session.add(chunk)
# Update processing session
processing_time = (time.time() - processing_start) * 1000
proc_session.completed_at = datetime.utcnow()
proc_session.status = 'completed'
proc_session.total_processing_time_ms = processing_time
proc_session.chunks_created = len(documents)
session.commit()
logger.info(f"Successfully saved {len(documents)} documents to database in {processing_time:.0f}ms")
return True
except Exception as e:
logger.error(f"Failed to save to database: {e}")
return False
def cleanup_old_data(self, retention_days: int = 30) -> None:
"""Clean up old processing sessions and orphaned data"""
try:
cutoff_date = datetime.utcnow() - timedelta(days=retention_days)
with self.get_session() as session:
# Clean up old processing sessions
old_sessions = session.query(ProcessingSession).filter(
ProcessingSession.started_at < cutoff_date
).delete()
# Clean up invalid cache entries
invalid_cache = session.query(SystemCache).filter(
SystemCache.is_valid == False
).delete()
session.commit()
logger.info(f"Cleaned up {old_sessions} old sessions and {invalid_cache} invalid cache entries")
except Exception as e:
logger.error(f"Cleanup failed: {e}")
def get_processing_history(self, limit: int = 10) -> List[Dict[str, Any]]:
"""Get recent processing session history"""
try:
with self.get_session() as session:
sessions = session.query(ProcessingSession).order_by(
ProcessingSession.started_at.desc()
).limit(limit).all()
return [
{
'session_id': s.session_id,
'started_at': s.started_at.isoformat(),
'completed_at': s.completed_at.isoformat() if s.completed_at else None,
'status': s.status,
'documents_processed': s.documents_processed,
'chunks_created': s.chunks_created,
'processing_time_ms': s.total_processing_time_ms,
'documents_per_second': s.documents_per_second
}
for s in sessions
]
except Exception as e:
logger.error(f"Failed to get processing history: {e}")
return []
def clear_database(self) -> bool:
"""Clear all data from database (for testing/reset)"""
try:
with self.get_session() as session:
session.query(DocumentChunk).delete()
session.query(Document).delete()
session.query(ProcessingSession).delete()
session.query(SystemCache).delete()
session.commit()
logger.info("Database cleared successfully")
return True
except Exception as e:
logger.error(f"Failed to clear database: {e}")
return False
def _hash_file(self, file_path: Path) -> str:
"""Generate hash of file content"""
try:
with open(file_path, 'rb') as f:
return hashlib.md5(f.read()).hexdigest()
except Exception as e:
logger.warning(f"Failed to hash file {file_path}: {e}")
return ""
def _hash_text(self, text: str) -> str:
"""Generate hash of text content"""
return hashlib.md5(text.encode('utf-8')).hexdigest()
def _hash_config(self, config: Dict[str, Any]) -> str:
"""Generate hash of configuration dictionary"""
try:
import json
# Convert config to string, handling any non-serializable objects
config_str = json.dumps(config, sort_keys=True, default=str)
return hashlib.md5(config_str.encode('utf-8')).hexdigest()
except Exception as e:
logger.warning(f"Config hash generation failed: {e}")
# Fallback to string representation
config_str = str(sorted(config.items()))
return hashlib.md5(config_str.encode('utf-8')).hexdigest()
# Global database manager instance
_db_manager = None
def get_database_manager(database_url: str = "sqlite:///demo/epic2_demo.db") -> DatabaseManager:
"""Get global database manager instance"""
global _db_manager
if _db_manager is None:
_db_manager = DatabaseManager(database_url)
return _db_manager
def reset_database_manager():
"""Reset global database manager (for testing)"""
global _db_manager
_db_manager = None |