Spaces:
Sleeping
Sleeping
File size: 22,398 Bytes
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 |
"""
Weaviate backend adapter for advanced retriever.
This module provides an adapter for Weaviate vector database integration,
following the same pattern as other external service adapters in the system.
It handles Weaviate-specific operations, schema management, and hybrid search.
"""
import logging
import time
import uuid
from datetime import datetime
from typing import List, Dict, Any, Optional, Tuple, Union
import numpy as np
try:
import weaviate
from weaviate import Client
from weaviate.exceptions import WeaviateException
WEAVIATE_AVAILABLE = True
except ImportError:
WEAVIATE_AVAILABLE = False
weaviate = None
Client = None
WeaviateException = Exception
from src.core.interfaces import Document
from .weaviate_config import WeaviateBackendConfig
logger = logging.getLogger(__name__)
class WeaviateConnectionError(Exception):
"""Raised when Weaviate connection fails."""
pass
class WeaviateSchemaError(Exception):
"""Raised when Weaviate schema operations fail."""
pass
class WeaviateBackend:
"""
Weaviate backend adapter for advanced retriever.
This adapter provides integration with Weaviate vector database,
following the adapter pattern used for external services like Ollama.
It handles schema management, document indexing, and hybrid search.
Features:
- Automatic schema creation and management
- Hybrid search (vector + keyword)
- Batch operations for performance
- Comprehensive error handling with fallbacks
- Performance monitoring and statistics
- Health checking and diagnostics
The adapter follows external service patterns:
- Connection management with retries
- Service-specific error handling
- Configuration-driven behavior
- Comprehensive logging and monitoring
"""
def __init__(self, config: Union[Dict[str, Any], WeaviateBackendConfig]):
"""
Initialize Weaviate backend adapter.
Args:
config: Configuration dictionary or WeaviateBackendConfig instance
Raises:
ImportError: If weaviate-client is not installed
WeaviateConnectionError: If connection fails
"""
if not WEAVIATE_AVAILABLE:
raise ImportError(
"weaviate-client is required for Weaviate backend. "
"Install with: pip install weaviate-client"
)
# Convert config if needed
if isinstance(config, dict):
self.config = WeaviateBackendConfig.from_dict(config)
else:
self.config = config
# Initialize client and connection
self.client: Optional[Client] = None
self.is_connected = False
self.schema_created = False
# Performance tracking
self.stats = {
"total_operations": 0,
"total_time": 0.0,
"avg_time": 0.0,
"last_operation_time": 0.0,
"search_count": 0,
"add_count": 0,
"error_count": 0,
"connection_errors": 0,
"schema_errors": 0
}
# Backend identification
self.backend_type = "weaviate"
self.backend_version = "adapter"
# Initialize connection
self._connect()
logger.info("Weaviate backend adapter initialized")
def _connect(self) -> None:
"""
Establish connection to Weaviate server.
Raises:
WeaviateConnectionError: If connection fails after retries
"""
if not WEAVIATE_AVAILABLE or weaviate is None:
raise WeaviateConnectionError("Weaviate package not available. Install with 'pip install weaviate-client'")
for attempt in range(self.config.max_retries + 1):
try:
# Create client with configuration
client_config = {
'url': self.config.connection.url,
'timeout_config': weaviate.TimeoutConfig(
query=self.config.connection.timeout,
insert=self.config.connection.timeout * 2
),
'startup_period': self.config.connection.startup_period
}
# Add authentication if provided
if self.config.connection.api_key:
client_config['auth_client_secret'] = weaviate.AuthApiKey(
api_key=self.config.connection.api_key
)
# Add additional headers
if self.config.connection.additional_headers:
client_config['additional_headers'] = self.config.connection.additional_headers
self.client = weaviate.Client(**client_config)
# Test connection
if self.client.is_ready():
self.is_connected = True
logger.info(f"Connected to Weaviate at {self.config.connection.url}")
# Create schema if needed
if self.config.auto_create_schema:
self._ensure_schema()
return
else:
raise WeaviateConnectionError("Weaviate server not ready")
except Exception as e:
self.stats["connection_errors"] += 1
logger.warning(f"Connection attempt {attempt + 1} failed: {str(e)}")
if attempt < self.config.max_retries:
time.sleep(self.config.retry_delay_seconds * (2 ** attempt)) # Exponential backoff
else:
raise WeaviateConnectionError(f"Failed to connect to Weaviate after {self.config.max_retries + 1} attempts: {str(e)}")
def _ensure_schema(self) -> None:
"""
Ensure the required schema exists in Weaviate.
Raises:
WeaviateSchemaError: If schema creation fails
"""
try:
# Check if class already exists
existing_schema = self.client.schema.get()
class_names = [cls['class'] for cls in existing_schema.get('classes', [])]
if self.config.schema.class_name in class_names:
logger.info(f"Schema class '{self.config.schema.class_name}' already exists")
self.schema_created = True
return
# Create new class
class_definition = {
"class": self.config.schema.class_name,
"description": self.config.schema.description,
"vectorizer": "none", # We provide our own embeddings
"vectorIndexConfig": self.config.schema.vector_index_config,
"properties": self.config.schema.properties
}
self.client.schema.create_class(class_definition)
self.schema_created = True
logger.info(f"Created schema class '{self.config.schema.class_name}'")
except Exception as e:
self.stats["schema_errors"] += 1
error_msg = f"Failed to create schema: {str(e)}"
logger.error(error_msg)
raise WeaviateSchemaError(error_msg) from e
def initialize_index(self, embedding_dim: int) -> None:
"""
Initialize the Weaviate index with specified dimension.
Args:
embedding_dim: Dimension of the embedding vectors
Note:
Weaviate handles dimensionality automatically, but we validate
that the connection is ready and schema exists.
"""
start_time = time.time()
try:
if not self.is_connected:
self._connect()
if not self.schema_created and self.config.auto_create_schema:
self._ensure_schema()
# Validate embedding dimension (informational)
logger.info(f"Weaviate backend ready for embeddings with dimension {embedding_dim}")
# Update stats
elapsed_time = time.time() - start_time
self._update_stats("initialize", elapsed_time)
except Exception as e:
self.stats["error_count"] += 1
logger.error(f"Failed to initialize Weaviate backend: {str(e)}")
raise RuntimeError(f"Weaviate backend initialization failed: {str(e)}") from e
def add_documents(self, documents: List[Document]) -> None:
"""
Add documents to the Weaviate index using batch operations.
Args:
documents: List of documents with embeddings to add
"""
start_time = time.time()
try:
if not documents:
raise ValueError("Cannot add empty document list")
if not self.is_connected:
self._connect()
# Validate embeddings
for i, doc in enumerate(documents):
if doc.embedding is None:
raise ValueError(f"Document {i} missing embedding")
# Batch insert
with self.client.batch(
batch_size=self.config.batch.batch_size,
num_workers=self.config.batch.num_workers,
connection_error_retries=self.config.batch.connection_error_retries,
timeout_retries=self.config.batch.timeout_retries,
callback=self._batch_callback if self.config.batch.callback_period else None
) as batch:
for i, document in enumerate(documents):
# Prepare document properties
properties = {
"content": document.content,
"source_file": document.metadata.get("source", "unknown"),
"chunk_index": document.metadata.get("chunk_index", i),
"page_number": document.metadata.get("page", 0),
"chunk_size": len(document.content),
"created_at": datetime.now().isoformat()
}
# Add additional metadata
for key, value in document.metadata.items():
if key not in properties and isinstance(value, (str, int, float, bool)):
properties[key] = value
# Generate UUID for document
doc_uuid = str(uuid.uuid4())
# Add to batch
batch.add_data_object(
data_object=properties,
class_name=self.config.schema.class_name,
uuid=doc_uuid,
vector=document.embedding
)
# Update stats
elapsed_time = time.time() - start_time
self._update_stats("add", elapsed_time)
self.stats["add_count"] += len(documents)
logger.info(f"Added {len(documents)} documents to Weaviate backend in {elapsed_time:.2f}s")
except Exception as e:
self.stats["error_count"] += 1
logger.error(f"Failed to add documents to Weaviate backend: {str(e)}")
raise RuntimeError(f"Weaviate backend add failed: {str(e)}") from e
def search(self, query_embedding: np.ndarray, k: int = 5, query_text: Optional[str] = None) -> List[Tuple[int, float]]:
"""
Search for similar documents using Weaviate.
Args:
query_embedding: Query vector
k: Number of results to return
query_text: Optional query text for hybrid search
Returns:
List of (document_index, score) tuples
"""
start_time = time.time()
try:
if k <= 0:
raise ValueError("k must be positive")
if not self.is_connected:
self._connect()
# Build query
query_builder = (
self.client.query
.get(self.config.schema.class_name, ["content", "source_file", "chunk_index"])
.with_limit(k)
)
# Use hybrid search if text query provided and enabled
if query_text and self.config.search.hybrid_search_enabled:
query_builder = query_builder.with_hybrid(
query=query_text,
alpha=self.config.search.alpha,
vector=query_embedding.tolist()
)
else:
# Pure vector search
query_builder = query_builder.with_near_vector({
"vector": query_embedding.tolist(),
"certainty": self.config.search.certainty_threshold
})
# Add additional search parameters
if self.config.search.autocut > 0:
query_builder = query_builder.with_autocut(self.config.search.autocut)
# Execute query
result = query_builder.do()
# Process results
results = []
if 'data' in result and 'Get' in result['data']:
class_results = result['data']['Get'].get(self.config.schema.class_name, [])
for i, item in enumerate(class_results):
# Extract score
if '_additional' in item:
if 'score' in item['_additional']:
score = float(item['_additional']['score'])
elif 'certainty' in item['_additional']:
score = float(item['_additional']['certainty'])
elif 'distance' in item['_additional']:
# Convert distance to similarity score
distance = float(item['_additional']['distance'])
score = 1.0 / (1.0 + distance)
else:
score = 1.0 - (i * 0.1) # Default decreasing score
else:
score = 1.0 - (i * 0.1) # Default decreasing score
# Use chunk_index if available, otherwise use position
doc_index = item.get('chunk_index', i)
results.append((doc_index, score))
# Update stats
elapsed_time = time.time() - start_time
self._update_stats("search", elapsed_time)
self.stats["search_count"] += 1
logger.debug(f"Weaviate search returned {len(results)} results in {elapsed_time:.4f}s")
return results
except Exception as e:
self.stats["error_count"] += 1
logger.error(f"Weaviate backend search failed: {str(e)}")
raise RuntimeError(f"Weaviate backend search failed: {str(e)}") from e
def get_document_count(self) -> int:
"""
Get the number of documents in the index.
Returns:
Number of indexed documents
"""
try:
if not self.is_connected:
return 0
result = (
self.client.query
.aggregate(self.config.schema.class_name)
.with_meta_count()
.do()
)
if 'data' in result and 'Aggregate' in result['data']:
aggregate_data = result['data']['Aggregate'].get(self.config.schema.class_name, [])
if aggregate_data:
return aggregate_data[0].get('meta', {}).get('count', 0)
return 0
except Exception as e:
logger.error(f"Failed to get document count: {str(e)}")
return 0
def is_trained(self) -> bool:
"""
Check if the index is ready for use.
Returns:
True if index is ready, False otherwise
"""
return self.is_connected and self.schema_created
def clear(self) -> None:
"""Clear all documents from the index."""
start_time = time.time()
try:
if not self.is_connected:
logger.warning("Not connected to Weaviate, cannot clear")
return
# Delete all objects of the class
self.client.batch.delete_objects(
class_name=self.config.schema.class_name,
where={
"operator": "Like",
"path": ["content"],
"valueText": "*"
}
)
# Update stats
elapsed_time = time.time() - start_time
self._update_stats("clear", elapsed_time)
logger.info("Weaviate backend cleared")
except Exception as e:
self.stats["error_count"] += 1
logger.error(f"Failed to clear Weaviate backend: {str(e)}")
raise RuntimeError(f"Weaviate backend clear failed: {str(e)}") from e
def get_backend_info(self) -> Dict[str, Any]:
"""
Get information about the backend.
Returns:
Dictionary with backend information
"""
try:
# Get Weaviate meta info
meta_info = {}
if self.is_connected:
try:
meta_info = self.client.get_meta()
except Exception:
meta_info = {"error": "Could not retrieve meta info"}
return {
"backend_type": self.backend_type,
"backend_version": self.backend_version,
"is_connected": self.is_connected,
"schema_created": self.schema_created,
"document_count": self.get_document_count(),
"weaviate_meta": meta_info,
"stats": self.stats.copy(),
"config": self.config.to_dict()
}
except Exception as e:
return {
"backend_type": self.backend_type,
"error": str(e),
"stats": self.stats.copy()
}
def health_check(self) -> Dict[str, Any]:
"""
Perform a health check on the backend.
Returns:
Dictionary with health status
"""
try:
is_healthy = True
issues = []
# Check connection
if not self.is_connected:
is_healthy = False
issues.append("Not connected to Weaviate")
elif self.client and not self.client.is_ready():
is_healthy = False
issues.append("Weaviate server not ready")
# Check schema
if not self.schema_created:
is_healthy = False
issues.append("Schema not created")
# Check error rate
error_rate = self.stats["error_count"] / max(1, self.stats["total_operations"])
if error_rate > 0.1: # More than 10% errors
is_healthy = False
issues.append(f"High error rate: {error_rate:.2%}")
# Check if we have documents
doc_count = self.get_document_count()
if doc_count == 0:
issues.append("No documents indexed")
return {
"backend_type": "weaviate",
"is_healthy": is_healthy,
"issues": issues,
"is_connected": self.is_connected,
"schema_created": self.schema_created,
"document_count": doc_count,
"error_rate": error_rate,
"total_operations": self.stats["total_operations"]
}
except Exception as e:
return {
"backend_type": "weaviate",
"is_healthy": False,
"issues": [f"Health check failed: {str(e)}"],
"error": str(e)
}
def supports_hybrid_search(self) -> bool:
"""
Check if backend supports hybrid search.
Returns:
True for Weaviate (supports vector + keyword search)
"""
return True
def supports_filtering(self) -> bool:
"""
Check if backend supports metadata filtering.
Returns:
True for Weaviate (supports where filters)
"""
return True
def _update_stats(self, operation: str, elapsed_time: float) -> None:
"""
Update performance statistics.
Args:
operation: Name of the operation
elapsed_time: Time taken for the operation
"""
self.stats["total_operations"] += 1
self.stats["total_time"] += elapsed_time
self.stats["avg_time"] = self.stats["total_time"] / self.stats["total_operations"]
self.stats["last_operation_time"] = elapsed_time
def _batch_callback(self, results: Dict[str, Any]) -> None:
"""
Callback for batch operations.
Args:
results: Batch operation results
"""
if results:
logger.debug(f"Batch operation completed: {len(results)} items processed")
def get_configuration(self) -> Dict[str, Any]:
"""
Get the current configuration.
Returns:
Configuration dictionary
"""
return {
"backend_type": "weaviate",
"config": self.config.to_dict()
} |