mirror of
https://github.com/trustgraph-ai/trustgraph.git
synced 2026-06-15 17:55:12 +02:00
a2dde9cafb
Cassandra triples services were using syncronous EntityCentricKnowledgeGraph methods from async contexts, and connection state was managed with threading.local which is wrong for asyncio coroutines sharing a single thread. Qdrant services had no async wrapping at all, blocking the event loop on every network call. Rows services had unprotected shared state mutations across concurrent coroutines. - Add async methods to EntityCentricKnowledgeGraph (async_insert, async_get_s/p/o/sp/po/os/spo/all, async_collection_exists, async_create_collection, async_delete_collection) using the existing cassandra_async.async_execute bridge - Rewrite triples write + query services: replace threading.local with asyncio.Lock + dict cache for per-workspace connections, use async ECKG methods for all data operations, keep asyncio.to_thread only for one-time blocking ECKG construction - Wrap all Qdrant calls in asyncio.to_thread across all 6 services (doc/graph/row embeddings write + query), add asyncio.Lock + set cache for collection existence checks - Add asyncio.Lock to rows write + query services to protect shared state (schemas, sessions, config caches) from concurrent mutation - Update all affected tests to match new async patterns
104 lines
2.6 KiB
Python
Executable file
104 lines
2.6 KiB
Python
Executable file
|
|
"""
|
|
Document embeddings query service. Input is vector, output is an array
|
|
of chunk_ids
|
|
"""
|
|
|
|
import asyncio
|
|
import logging
|
|
|
|
from qdrant_client import QdrantClient
|
|
|
|
from .... schema import DocumentEmbeddingsResponse, ChunkMatch
|
|
from .... schema import Error
|
|
from .... base import DocumentEmbeddingsQueryService
|
|
|
|
# Module logger
|
|
logger = logging.getLogger(__name__)
|
|
|
|
default_ident = "doc-embeddings-query"
|
|
|
|
default_store_uri = 'http://localhost:6333'
|
|
|
|
class Processor(DocumentEmbeddingsQueryService):
|
|
|
|
def __init__(self, **params):
|
|
|
|
store_uri = params.get("store_uri", default_store_uri)
|
|
|
|
#optional api key
|
|
api_key = params.get("api_key", None)
|
|
|
|
super(Processor, self).__init__(
|
|
**params | {
|
|
"store_uri": store_uri,
|
|
"api_key": api_key,
|
|
}
|
|
)
|
|
|
|
self.qdrant = QdrantClient(url=store_uri, api_key=api_key)
|
|
|
|
async def query_document_embeddings(self, workspace, msg):
|
|
|
|
try:
|
|
|
|
vec = msg.vector
|
|
if not vec:
|
|
return []
|
|
|
|
dim = len(vec)
|
|
collection = f"d_{workspace}_{msg.collection}_{dim}"
|
|
|
|
exists = await asyncio.to_thread(
|
|
self.qdrant.collection_exists, collection
|
|
)
|
|
if not exists:
|
|
logger.info(f"Collection {collection} does not exist, returning empty results")
|
|
return []
|
|
|
|
result = await asyncio.to_thread(
|
|
self.qdrant.query_points,
|
|
collection_name=collection,
|
|
query=vec,
|
|
limit=msg.limit,
|
|
with_payload=True,
|
|
)
|
|
search_result = result.points
|
|
|
|
chunks = []
|
|
for r in search_result:
|
|
chunk_id = r.payload["chunk_id"]
|
|
score = r.score if hasattr(r, 'score') else 0.0
|
|
chunks.append(ChunkMatch(
|
|
chunk_id=chunk_id,
|
|
score=score,
|
|
))
|
|
|
|
return chunks
|
|
|
|
except Exception as e:
|
|
|
|
logger.error(f"Exception querying document embeddings: {e}", exc_info=True)
|
|
raise e
|
|
|
|
@staticmethod
|
|
def add_args(parser):
|
|
|
|
DocumentEmbeddingsQueryService.add_args(parser)
|
|
|
|
parser.add_argument(
|
|
'-t', '--store-uri',
|
|
default=default_store_uri,
|
|
help=f'Qdrant store URI (default: {default_store_uri})'
|
|
)
|
|
|
|
parser.add_argument(
|
|
'-k', '--api-key',
|
|
default=None,
|
|
help=f'API key for qdrant (default: None)'
|
|
)
|
|
|
|
def run():
|
|
|
|
Processor.launch(default_ident, __doc__)
|
|
|