File size: 14,764 Bytes
cfb0fa4 | 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 | from typing import Optional, List, Dict, Any
import logging
import re
import json
from sqlalchemy import (
func,
literal,
cast,
column,
create_engine,
Column,
Integer,
MetaData,
LargeBinary,
select,
text,
Text,
Table,
values,
)
from sqlalchemy.sql import true
from sqlalchemy.pool import NullPool, QueuePool
from sqlalchemy.orm import declarative_base, scoped_session, sessionmaker
from sqlalchemy.dialects.postgresql import JSONB, array
from pgvector.sqlalchemy import Vector
from sqlalchemy.ext.mutable import MutableDict
from sqlalchemy.exc import NoSuchTableError
from sqlalchemy.dialects.postgresql.psycopg2 import PGDialect_psycopg2
from sqlalchemy.dialects import registry
class OpenGaussDialect(PGDialect_psycopg2):
name = "opengauss"
def _get_server_version_info(self, connection):
try:
version = connection.exec_driver_sql("SELECT version()").scalar()
if not version:
return (9, 0, 0)
match = re.search(
r"openGauss\s+(\d+)\.(\d+)\.(\d+)(?:-\w+)?", version, re.IGNORECASE
)
if match:
return (int(match.group(1)), int(match.group(2)), int(match.group(3)))
return super()._get_server_version_info(connection)
except Exception:
return (9, 0, 0)
# Register dialect
registry.register("opengauss", __name__, "OpenGaussDialect")
from open_webui.retrieval.vector.utils import process_metadata
from open_webui.retrieval.vector.main import (
VectorDBBase,
VectorItem,
SearchResult,
GetResult,
)
from open_webui.config import (
OPENGAUSS_DB_URL,
OPENGAUSS_INITIALIZE_MAX_VECTOR_LENGTH,
OPENGAUSS_POOL_SIZE,
OPENGAUSS_POOL_MAX_OVERFLOW,
OPENGAUSS_POOL_TIMEOUT,
OPENGAUSS_POOL_RECYCLE,
)
from open_webui.env import SRC_LOG_LEVELS
VECTOR_LENGTH = OPENGAUSS_INITIALIZE_MAX_VECTOR_LENGTH
Base = declarative_base()
log = logging.getLogger(__name__)
log.setLevel(SRC_LOG_LEVELS["RAG"])
class DocumentChunk(Base):
__tablename__ = "document_chunk"
id = Column(Text, primary_key=True)
vector = Column(Vector(dim=VECTOR_LENGTH), nullable=True)
collection_name = Column(Text, nullable=False)
text = Column(Text, nullable=True)
vmetadata = Column(MutableDict.as_mutable(JSONB), nullable=True)
class OpenGaussClient(VectorDBBase):
def __init__(self) -> None:
if not OPENGAUSS_DB_URL:
from open_webui.internal.db import ScopedSession
self.session = ScopedSession
else:
engine_kwargs = {"pool_pre_ping": True, "dialect": OpenGaussDialect()}
if isinstance(OPENGAUSS_POOL_SIZE, int) and OPENGAUSS_POOL_SIZE > 0:
engine_kwargs.update(
{
"pool_size": OPENGAUSS_POOL_SIZE,
"max_overflow": OPENGAUSS_POOL_MAX_OVERFLOW,
"pool_timeout": OPENGAUSS_POOL_TIMEOUT,
"pool_recycle": OPENGAUSS_POOL_RECYCLE,
"poolclass": QueuePool,
}
)
else:
engine_kwargs["poolclass"] = NullPool
engine = create_engine(OPENGAUSS_DB_URL, **engine_kwargs)
SessionLocal = sessionmaker(
autocommit=False, autoflush=False, bind=engine, expire_on_commit=False
)
self.session = scoped_session(SessionLocal)
try:
connection = self.session.connection()
Base.metadata.create_all(bind=connection)
self.session.execute(
text(
"CREATE INDEX IF NOT EXISTS idx_document_chunk_vector "
"ON document_chunk USING ivfflat (vector vector_cosine_ops) WITH (lists = 100);"
)
)
self.session.execute(
text(
"CREATE INDEX IF NOT EXISTS idx_document_chunk_collection_name "
"ON document_chunk (collection_name);"
)
)
self.session.commit()
log.info("OpenGauss vector database initialization completed.")
except Exception as e:
self.session.rollback()
log.exception(f"OpenGauss Initialization failed.: {e}")
raise
def check_vector_length(self) -> None:
metadata = MetaData()
try:
document_chunk_table = Table(
"document_chunk", metadata, autoload_with=self.session.bind
)
except NoSuchTableError:
return
if "vector" in document_chunk_table.columns:
vector_column = document_chunk_table.columns["vector"]
vector_type = vector_column.type
if isinstance(vector_type, Vector):
db_vector_length = vector_type.dim
if db_vector_length != VECTOR_LENGTH:
raise Exception(
f"Vector dimension mismatch: configured {VECTOR_LENGTH} vs. {db_vector_length} in the database."
)
else:
raise Exception("The 'vector' column type is not Vector.")
else:
raise Exception(
"The 'vector' column does not exist in the 'document_chunk' table."
)
def adjust_vector_length(self, vector: List[float]) -> List[float]:
current_length = len(vector)
if current_length < VECTOR_LENGTH:
vector += [0.0] * (VECTOR_LENGTH - current_length)
elif current_length > VECTOR_LENGTH:
vector = vector[:VECTOR_LENGTH]
return vector
def insert(self, collection_name: str, items: List[VectorItem]) -> None:
try:
new_items = []
for item in items:
vector = self.adjust_vector_length(item["vector"])
new_chunk = DocumentChunk(
id=item["id"],
vector=vector,
collection_name=collection_name,
text=item["text"],
vmetadata=process_metadata(item["metadata"]),
)
new_items.append(new_chunk)
self.session.bulk_save_objects(new_items)
self.session.commit()
log.info(
f"Inserting {len(new_items)} items into collection '{collection_name}'."
)
except Exception as e:
self.session.rollback()
log.exception(f"Failed to insert data: {e}")
raise
def upsert(self, collection_name: str, items: List[VectorItem]) -> None:
try:
for item in items:
vector = self.adjust_vector_length(item["vector"])
existing = (
self.session.query(DocumentChunk)
.filter(DocumentChunk.id == item["id"])
.first()
)
if existing:
existing.vector = vector
existing.text = item["text"]
existing.vmetadata = process_metadata(item["metadata"])
existing.collection_name = collection_name
else:
new_chunk = DocumentChunk(
id=item["id"],
vector=vector,
collection_name=collection_name,
text=item["text"],
vmetadata=process_metadata(item["metadata"]),
)
self.session.add(new_chunk)
self.session.commit()
log.info(
f"Inserting/updating {len(items)} items in collection '{collection_name}'."
)
except Exception as e:
self.session.rollback()
log.exception(f"Failed to insert or update data.: {e}")
raise
def search(
self,
collection_name: str,
vectors: List[List[float]],
filter: Optional[Dict[str, Any]] = None,
limit: int = 10,
) -> Optional[SearchResult]:
try:
if not vectors:
return None
vectors = [self.adjust_vector_length(vector) for vector in vectors]
num_queries = len(vectors)
def vector_expr(vector):
return cast(array(vector), Vector(VECTOR_LENGTH))
qid_col = column("qid", Integer)
q_vector_col = column("q_vector", Vector(VECTOR_LENGTH))
query_vectors = (
values(qid_col, q_vector_col)
.data(
[(idx, vector_expr(vector)) for idx, vector in enumerate(vectors)]
)
.alias("query_vectors")
)
result_fields = [
DocumentChunk.id,
DocumentChunk.text,
DocumentChunk.vmetadata,
(DocumentChunk.vector.cosine_distance(query_vectors.c.q_vector)).label(
"distance"
),
]
subq = (
select(*result_fields)
.where(DocumentChunk.collection_name == collection_name)
.order_by(
DocumentChunk.vector.cosine_distance(query_vectors.c.q_vector)
)
)
if limit is not None:
subq = subq.limit(limit)
subq = subq.lateral("result")
stmt = (
select(
query_vectors.c.qid,
subq.c.id,
subq.c.text,
subq.c.vmetadata,
subq.c.distance,
)
.select_from(query_vectors)
.join(subq, true())
.order_by(query_vectors.c.qid, subq.c.distance)
)
result_proxy = self.session.execute(stmt)
results = result_proxy.all()
ids = [[] for _ in range(num_queries)]
distances = [[] for _ in range(num_queries)]
documents = [[] for _ in range(num_queries)]
metadatas = [[] for _ in range(num_queries)]
for row in results:
qid = int(row.qid)
ids[qid].append(row.id)
distances[qid].append((2.0 - row.distance) / 2.0)
documents[qid].append(row.text)
metadatas[qid].append(row.vmetadata)
self.session.rollback()
return SearchResult(
ids=ids, distances=distances, documents=documents, metadatas=metadatas
)
except Exception as e:
self.session.rollback()
log.exception(f"Vector search failed: {e}")
return None
def query(
self, collection_name: str, filter: Dict[str, Any], limit: Optional[int] = None
) -> Optional[GetResult]:
try:
query = self.session.query(DocumentChunk).filter(
DocumentChunk.collection_name == collection_name
)
for key, value in filter.items():
query = query.filter(DocumentChunk.vmetadata[key].astext == str(value))
if limit is not None:
query = query.limit(limit)
results = query.all()
if not results:
return None
ids = [[result.id for result in results]]
documents = [[result.text for result in results]]
metadatas = [[result.vmetadata for result in results]]
self.session.rollback()
return GetResult(ids=ids, documents=documents, metadatas=metadatas)
except Exception as e:
self.session.rollback()
log.exception(f"Conditional query failed: {e}")
return None
def get(
self, collection_name: str, limit: Optional[int] = None
) -> Optional[GetResult]:
try:
query = self.session.query(DocumentChunk).filter(
DocumentChunk.collection_name == collection_name
)
if limit is not None:
query = query.limit(limit)
results = query.all()
if not results:
return None
ids = [[result.id for result in results]]
documents = [[result.text for result in results]]
metadatas = [[result.vmetadata for result in results]]
self.session.rollback()
return GetResult(ids=ids, documents=documents, metadatas=metadatas)
except Exception as e:
self.session.rollback()
log.exception(f"Failed to retrieve data: {e}")
return None
def delete(
self,
collection_name: str,
ids: Optional[List[str]] = None,
filter: Optional[Dict[str, Any]] = None,
) -> None:
try:
query = self.session.query(DocumentChunk).filter(
DocumentChunk.collection_name == collection_name
)
if ids:
query = query.filter(DocumentChunk.id.in_(ids))
if filter:
for key, value in filter.items():
query = query.filter(
DocumentChunk.vmetadata[key].astext == str(value)
)
deleted = query.delete(synchronize_session=False)
self.session.commit()
log.info(f"Deleted {deleted} items from collection '{collection_name}'")
except Exception as e:
self.session.rollback()
log.exception(f"Failed to delete data: {e}")
raise
def reset(self) -> None:
try:
deleted = self.session.query(DocumentChunk).delete()
self.session.commit()
log.info(f"Reset completed. Deleted {deleted} items")
except Exception as e:
self.session.rollback()
log.exception(f"Reset failed: {e}")
raise
def close(self) -> None:
pass
def has_collection(self, collection_name: str) -> bool:
try:
exists = (
self.session.query(DocumentChunk)
.filter(DocumentChunk.collection_name == collection_name)
.first()
is not None
)
self.session.rollback()
return exists
except Exception as e:
self.session.rollback()
log.exception(f"Failed to check collection existence: {e}")
return False
def delete_collection(self, collection_name: str) -> None:
self.delete(collection_name)
log.info(f"Collection '{collection_name}' has been deleted")
|