trustgraph/tests/unit/test_reliability/test_null_embedding_protection.py

315 lines
11 KiB
Python
Raw Normal View History

"""
Tests for null embedding protection: empty/None vector skipping, entity
validation, dimension-aware collection creation, and query-time empty
vector handling.
Tests the pure functions and logic without Qdrant connections.
"""
import pytest
from unittest.mock import MagicMock, patch, AsyncMock
from trustgraph.schema import Term, IRI, LITERAL, BLANK
# ---------------------------------------------------------------------------
# Graph embeddings: get_term_value
# ---------------------------------------------------------------------------
class TestGraphEmbeddingsGetTermValue:
def test_iri_returns_iri(self):
from trustgraph.storage.graph_embeddings.qdrant.write import get_term_value
t = Term(type=IRI, iri="http://example.org/x")
assert get_term_value(t) == "http://example.org/x"
def test_literal_returns_value(self):
from trustgraph.storage.graph_embeddings.qdrant.write import get_term_value
t = Term(type=LITERAL, value="hello")
assert get_term_value(t) == "hello"
def test_blank_returns_id(self):
from trustgraph.storage.graph_embeddings.qdrant.write import get_term_value
t = Term(type=BLANK, id="_:b0")
assert get_term_value(t) == "_:b0"
def test_none_returns_none(self):
from trustgraph.storage.graph_embeddings.qdrant.write import get_term_value
assert get_term_value(None) is None
def test_blank_with_value_fallback(self):
from trustgraph.storage.graph_embeddings.qdrant.write import get_term_value
t = Term(type=BLANK, id="", value="fallback")
assert get_term_value(t) == "fallback"
# ---------------------------------------------------------------------------
# Document embeddings: null vector protection
# ---------------------------------------------------------------------------
class TestDocEmbeddingsNullProtection:
@pytest.mark.asyncio
async def test_empty_vector_skipped(self):
"""Embeddings with empty vectors should be silently skipped."""
from trustgraph.storage.doc_embeddings.qdrant.write import Processor
proc = Processor.__new__(Processor)
proc.qdrant = MagicMock()
# Mock collection_exists for config check
proc.collection_exists = MagicMock(return_value=True)
msg = MagicMock()
msg.metadata.collection = "col1"
emb = MagicMock()
emb.chunk_id = "chunk-1"
emb.vector = [] # Empty vector
msg.chunks = [emb]
2026-04-22 15:19:57 +01:00
await proc.store_document_embeddings("user1", msg)
# No upsert should be called
proc.qdrant.upsert.assert_not_called()
@pytest.mark.asyncio
async def test_none_vector_skipped(self):
from trustgraph.storage.doc_embeddings.qdrant.write import Processor
proc = Processor.__new__(Processor)
proc.qdrant = MagicMock()
proc.collection_exists = MagicMock(return_value=True)
msg = MagicMock()
msg.metadata.collection = "col1"
emb = MagicMock()
emb.chunk_id = "chunk-1"
emb.vector = None # None vector
msg.chunks = [emb]
2026-04-22 15:19:57 +01:00
await proc.store_document_embeddings("user1", msg)
proc.qdrant.upsert.assert_not_called()
@pytest.mark.asyncio
async def test_empty_chunk_id_skipped(self):
from trustgraph.storage.doc_embeddings.qdrant.write import Processor
proc = Processor.__new__(Processor)
proc.qdrant = MagicMock()
proc.collection_exists = MagicMock(return_value=True)
msg = MagicMock()
msg.metadata.collection = "col1"
emb = MagicMock()
emb.chunk_id = "" # Empty chunk ID
emb.vector = [0.1, 0.2, 0.3]
msg.chunks = [emb]
2026-04-22 15:19:57 +01:00
await proc.store_document_embeddings("user1", msg)
proc.qdrant.upsert.assert_not_called()
@pytest.mark.asyncio
async def test_valid_embedding_upserted(self):
release/v2.4 -> master (#924) * CLI auth migration, document embeddings core lifecycle (#913) Migrate get_kg_core and put_kg_core CLI tools to use Api/SocketClient with first-frame auth (fixes broken raw websocket path). Fix wire format field names (root/vector). Remove ~600 lines of dead raw websocket code from invoke_graph_rag.py. Add document embeddings core lifecycle to the knowledge service: list/get/put/delete/load operations across schema, translator, Cassandra table store, knowledge manager, gateway registry, REST API, socket client, and CLI (tg-get-de-core, tg-put-de-core). Fix delete_kg_core to also clean up document embeddings rows. * Remove spurious workspace parameter from SPARQL algebra evaluator (#915) Fix threading of workspace paramater: - The SPARQL algebra evaluator was threading a workspace parameter through every function and passing it to TriplesClient.query(), which doesn't accept it. Workspace isolation is handled by pub/sub topic routing — the TriplesClient is already scoped to a workspace-specific flow, same as GraphRAG. Passing workspace explicitly was both incorrect and unnecessary. Update tests: - tests/unit/test_query/test_sparql_algebra.py (new) — Tests _query_pattern, _eval_bgp, and evaluate() with various algebra nodes. Key tests assert workspace is never in tc.query() kwargs, plus correctness tests for BGP, JOIN, UNION, SLICE, DISTINCT, and edge cases. - tests/unit/test_retrieval/test_graph_rag.py — Added test_triples_query_never_passes_workspace (checks query()) and test_follow_edges_never_passes_workspace (checks query_stream()). * Make all Cassandra and Qdrant I/O async-safe with proper concurrency controls (#916) 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 * Fixed error only returning a page of results (#921) The root cause: async_execute only materialises the first result page (by design — it says so in its docstring). The streaming query set fetch_size=20 and expected to iterate all results, but only got the first 20 rows back. The fix uses asyncio.to_thread(lambda: list(tg.session.execute(...))) which lets the sync driver iterate all pages in a worker thread — exactly what the pre-async code did. * Optional test warning suppression (#923) * Fix test collection module errors & silence upstream Pytest warnings (#823) * chore: add virtual environment and .env directories to gitignore * test: filter upstream DeprecationWarning and UserWarning messages * fix(namespace): remove empty __init__.py files to fix PEP 420 implicit namespace routing for trustgraph sub-packages * Revert __init__.py deletions * Add .ini changes but commented out, will be useful at times --------- Co-authored-by: Salil M <d2kyt@protonmail.com>
2026-05-15 13:02:51 +01:00
import asyncio
from trustgraph.storage.doc_embeddings.qdrant.write import Processor
proc = Processor.__new__(Processor)
proc.qdrant = MagicMock()
proc.qdrant.collection_exists.return_value = True
proc.collection_exists = MagicMock(return_value=True)
release/v2.4 -> master (#924) * CLI auth migration, document embeddings core lifecycle (#913) Migrate get_kg_core and put_kg_core CLI tools to use Api/SocketClient with first-frame auth (fixes broken raw websocket path). Fix wire format field names (root/vector). Remove ~600 lines of dead raw websocket code from invoke_graph_rag.py. Add document embeddings core lifecycle to the knowledge service: list/get/put/delete/load operations across schema, translator, Cassandra table store, knowledge manager, gateway registry, REST API, socket client, and CLI (tg-get-de-core, tg-put-de-core). Fix delete_kg_core to also clean up document embeddings rows. * Remove spurious workspace parameter from SPARQL algebra evaluator (#915) Fix threading of workspace paramater: - The SPARQL algebra evaluator was threading a workspace parameter through every function and passing it to TriplesClient.query(), which doesn't accept it. Workspace isolation is handled by pub/sub topic routing — the TriplesClient is already scoped to a workspace-specific flow, same as GraphRAG. Passing workspace explicitly was both incorrect and unnecessary. Update tests: - tests/unit/test_query/test_sparql_algebra.py (new) — Tests _query_pattern, _eval_bgp, and evaluate() with various algebra nodes. Key tests assert workspace is never in tc.query() kwargs, plus correctness tests for BGP, JOIN, UNION, SLICE, DISTINCT, and edge cases. - tests/unit/test_retrieval/test_graph_rag.py — Added test_triples_query_never_passes_workspace (checks query()) and test_follow_edges_never_passes_workspace (checks query_stream()). * Make all Cassandra and Qdrant I/O async-safe with proper concurrency controls (#916) 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 * Fixed error only returning a page of results (#921) The root cause: async_execute only materialises the first result page (by design — it says so in its docstring). The streaming query set fetch_size=20 and expected to iterate all results, but only got the first 20 rows back. The fix uses asyncio.to_thread(lambda: list(tg.session.execute(...))) which lets the sync driver iterate all pages in a worker thread — exactly what the pre-async code did. * Optional test warning suppression (#923) * Fix test collection module errors & silence upstream Pytest warnings (#823) * chore: add virtual environment and .env directories to gitignore * test: filter upstream DeprecationWarning and UserWarning messages * fix(namespace): remove empty __init__.py files to fix PEP 420 implicit namespace routing for trustgraph sub-packages * Revert __init__.py deletions * Add .ini changes but commented out, will be useful at times --------- Co-authored-by: Salil M <d2kyt@protonmail.com>
2026-05-15 13:02:51 +01:00
proc._cache_lock = asyncio.Lock()
proc._known_collections = set()
msg = MagicMock()
msg.metadata.collection = "col1"
emb = MagicMock()
emb.chunk_id = "chunk-1"
emb.vector = [0.1, 0.2, 0.3]
msg.chunks = [emb]
2026-04-22 15:19:57 +01:00
await proc.store_document_embeddings("user1", msg)
proc.qdrant.upsert.assert_called_once()
@pytest.mark.asyncio
async def test_dimension_in_collection_name(self):
"""Collection name should include vector dimension."""
release/v2.4 -> master (#924) * CLI auth migration, document embeddings core lifecycle (#913) Migrate get_kg_core and put_kg_core CLI tools to use Api/SocketClient with first-frame auth (fixes broken raw websocket path). Fix wire format field names (root/vector). Remove ~600 lines of dead raw websocket code from invoke_graph_rag.py. Add document embeddings core lifecycle to the knowledge service: list/get/put/delete/load operations across schema, translator, Cassandra table store, knowledge manager, gateway registry, REST API, socket client, and CLI (tg-get-de-core, tg-put-de-core). Fix delete_kg_core to also clean up document embeddings rows. * Remove spurious workspace parameter from SPARQL algebra evaluator (#915) Fix threading of workspace paramater: - The SPARQL algebra evaluator was threading a workspace parameter through every function and passing it to TriplesClient.query(), which doesn't accept it. Workspace isolation is handled by pub/sub topic routing — the TriplesClient is already scoped to a workspace-specific flow, same as GraphRAG. Passing workspace explicitly was both incorrect and unnecessary. Update tests: - tests/unit/test_query/test_sparql_algebra.py (new) — Tests _query_pattern, _eval_bgp, and evaluate() with various algebra nodes. Key tests assert workspace is never in tc.query() kwargs, plus correctness tests for BGP, JOIN, UNION, SLICE, DISTINCT, and edge cases. - tests/unit/test_retrieval/test_graph_rag.py — Added test_triples_query_never_passes_workspace (checks query()) and test_follow_edges_never_passes_workspace (checks query_stream()). * Make all Cassandra and Qdrant I/O async-safe with proper concurrency controls (#916) 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 * Fixed error only returning a page of results (#921) The root cause: async_execute only materialises the first result page (by design — it says so in its docstring). The streaming query set fetch_size=20 and expected to iterate all results, but only got the first 20 rows back. The fix uses asyncio.to_thread(lambda: list(tg.session.execute(...))) which lets the sync driver iterate all pages in a worker thread — exactly what the pre-async code did. * Optional test warning suppression (#923) * Fix test collection module errors & silence upstream Pytest warnings (#823) * chore: add virtual environment and .env directories to gitignore * test: filter upstream DeprecationWarning and UserWarning messages * fix(namespace): remove empty __init__.py files to fix PEP 420 implicit namespace routing for trustgraph sub-packages * Revert __init__.py deletions * Add .ini changes but commented out, will be useful at times --------- Co-authored-by: Salil M <d2kyt@protonmail.com>
2026-05-15 13:02:51 +01:00
import asyncio
from trustgraph.storage.doc_embeddings.qdrant.write import Processor
proc = Processor.__new__(Processor)
proc.qdrant = MagicMock()
proc.qdrant.collection_exists.return_value = True
proc.collection_exists = MagicMock(return_value=True)
release/v2.4 -> master (#924) * CLI auth migration, document embeddings core lifecycle (#913) Migrate get_kg_core and put_kg_core CLI tools to use Api/SocketClient with first-frame auth (fixes broken raw websocket path). Fix wire format field names (root/vector). Remove ~600 lines of dead raw websocket code from invoke_graph_rag.py. Add document embeddings core lifecycle to the knowledge service: list/get/put/delete/load operations across schema, translator, Cassandra table store, knowledge manager, gateway registry, REST API, socket client, and CLI (tg-get-de-core, tg-put-de-core). Fix delete_kg_core to also clean up document embeddings rows. * Remove spurious workspace parameter from SPARQL algebra evaluator (#915) Fix threading of workspace paramater: - The SPARQL algebra evaluator was threading a workspace parameter through every function and passing it to TriplesClient.query(), which doesn't accept it. Workspace isolation is handled by pub/sub topic routing — the TriplesClient is already scoped to a workspace-specific flow, same as GraphRAG. Passing workspace explicitly was both incorrect and unnecessary. Update tests: - tests/unit/test_query/test_sparql_algebra.py (new) — Tests _query_pattern, _eval_bgp, and evaluate() with various algebra nodes. Key tests assert workspace is never in tc.query() kwargs, plus correctness tests for BGP, JOIN, UNION, SLICE, DISTINCT, and edge cases. - tests/unit/test_retrieval/test_graph_rag.py — Added test_triples_query_never_passes_workspace (checks query()) and test_follow_edges_never_passes_workspace (checks query_stream()). * Make all Cassandra and Qdrant I/O async-safe with proper concurrency controls (#916) 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 * Fixed error only returning a page of results (#921) The root cause: async_execute only materialises the first result page (by design — it says so in its docstring). The streaming query set fetch_size=20 and expected to iterate all results, but only got the first 20 rows back. The fix uses asyncio.to_thread(lambda: list(tg.session.execute(...))) which lets the sync driver iterate all pages in a worker thread — exactly what the pre-async code did. * Optional test warning suppression (#923) * Fix test collection module errors & silence upstream Pytest warnings (#823) * chore: add virtual environment and .env directories to gitignore * test: filter upstream DeprecationWarning and UserWarning messages * fix(namespace): remove empty __init__.py files to fix PEP 420 implicit namespace routing for trustgraph sub-packages * Revert __init__.py deletions * Add .ini changes but commented out, will be useful at times --------- Co-authored-by: Salil M <d2kyt@protonmail.com>
2026-05-15 13:02:51 +01:00
proc._cache_lock = asyncio.Lock()
proc._known_collections = set()
msg = MagicMock()
msg.metadata.collection = "docs"
emb = MagicMock()
emb.chunk_id = "c1"
emb.vector = [0.0] * 384 # 384-dim vector
msg.chunks = [emb]
2026-04-22 15:19:57 +01:00
await proc.store_document_embeddings("alice", msg)
call_args = proc.qdrant.upsert.call_args
assert "d_alice_docs_384" in call_args[1]["collection_name"]
# ---------------------------------------------------------------------------
# Graph embeddings: null entity and vector protection
# ---------------------------------------------------------------------------
class TestGraphEmbeddingsNullProtection:
@pytest.mark.asyncio
async def test_empty_entity_skipped(self):
from trustgraph.storage.graph_embeddings.qdrant.write import Processor
proc = Processor.__new__(Processor)
proc.qdrant = MagicMock()
proc.collection_exists = MagicMock(return_value=True)
msg = MagicMock()
msg.metadata.collection = "col1"
entity = MagicMock()
entity.entity = Term(type=IRI, iri="") # Empty IRI
entity.vector = [0.1, 0.2, 0.3]
msg.entities = [entity]
2026-04-22 15:19:57 +01:00
await proc.store_graph_embeddings("user1", msg)
proc.qdrant.upsert.assert_not_called()
@pytest.mark.asyncio
async def test_none_entity_skipped(self):
from trustgraph.storage.graph_embeddings.qdrant.write import Processor
proc = Processor.__new__(Processor)
proc.qdrant = MagicMock()
proc.collection_exists = MagicMock(return_value=True)
msg = MagicMock()
msg.metadata.collection = "col1"
entity = MagicMock()
entity.entity = None # Null entity
entity.vector = [0.1, 0.2, 0.3]
msg.entities = [entity]
2026-04-22 15:19:57 +01:00
await proc.store_graph_embeddings("user1", msg)
proc.qdrant.upsert.assert_not_called()
@pytest.mark.asyncio
async def test_empty_vector_skipped(self):
from trustgraph.storage.graph_embeddings.qdrant.write import Processor
proc = Processor.__new__(Processor)
proc.qdrant = MagicMock()
proc.collection_exists = MagicMock(return_value=True)
msg = MagicMock()
msg.metadata.collection = "col1"
entity = MagicMock()
entity.entity = Term(type=IRI, iri="http://example.org/x")
entity.vector = [] # Empty vector
msg.entities = [entity]
2026-04-22 15:19:57 +01:00
await proc.store_graph_embeddings("user1", msg)
proc.qdrant.upsert.assert_not_called()
@pytest.mark.asyncio
async def test_valid_entity_and_vector_upserted(self):
release/v2.4 -> master (#924) * CLI auth migration, document embeddings core lifecycle (#913) Migrate get_kg_core and put_kg_core CLI tools to use Api/SocketClient with first-frame auth (fixes broken raw websocket path). Fix wire format field names (root/vector). Remove ~600 lines of dead raw websocket code from invoke_graph_rag.py. Add document embeddings core lifecycle to the knowledge service: list/get/put/delete/load operations across schema, translator, Cassandra table store, knowledge manager, gateway registry, REST API, socket client, and CLI (tg-get-de-core, tg-put-de-core). Fix delete_kg_core to also clean up document embeddings rows. * Remove spurious workspace parameter from SPARQL algebra evaluator (#915) Fix threading of workspace paramater: - The SPARQL algebra evaluator was threading a workspace parameter through every function and passing it to TriplesClient.query(), which doesn't accept it. Workspace isolation is handled by pub/sub topic routing — the TriplesClient is already scoped to a workspace-specific flow, same as GraphRAG. Passing workspace explicitly was both incorrect and unnecessary. Update tests: - tests/unit/test_query/test_sparql_algebra.py (new) — Tests _query_pattern, _eval_bgp, and evaluate() with various algebra nodes. Key tests assert workspace is never in tc.query() kwargs, plus correctness tests for BGP, JOIN, UNION, SLICE, DISTINCT, and edge cases. - tests/unit/test_retrieval/test_graph_rag.py — Added test_triples_query_never_passes_workspace (checks query()) and test_follow_edges_never_passes_workspace (checks query_stream()). * Make all Cassandra and Qdrant I/O async-safe with proper concurrency controls (#916) 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 * Fixed error only returning a page of results (#921) The root cause: async_execute only materialises the first result page (by design — it says so in its docstring). The streaming query set fetch_size=20 and expected to iterate all results, but only got the first 20 rows back. The fix uses asyncio.to_thread(lambda: list(tg.session.execute(...))) which lets the sync driver iterate all pages in a worker thread — exactly what the pre-async code did. * Optional test warning suppression (#923) * Fix test collection module errors & silence upstream Pytest warnings (#823) * chore: add virtual environment and .env directories to gitignore * test: filter upstream DeprecationWarning and UserWarning messages * fix(namespace): remove empty __init__.py files to fix PEP 420 implicit namespace routing for trustgraph sub-packages * Revert __init__.py deletions * Add .ini changes but commented out, will be useful at times --------- Co-authored-by: Salil M <d2kyt@protonmail.com>
2026-05-15 13:02:51 +01:00
import asyncio
from trustgraph.storage.graph_embeddings.qdrant.write import Processor
proc = Processor.__new__(Processor)
proc.qdrant = MagicMock()
proc.qdrant.collection_exists.return_value = True
proc.collection_exists = MagicMock(return_value=True)
release/v2.4 -> master (#924) * CLI auth migration, document embeddings core lifecycle (#913) Migrate get_kg_core and put_kg_core CLI tools to use Api/SocketClient with first-frame auth (fixes broken raw websocket path). Fix wire format field names (root/vector). Remove ~600 lines of dead raw websocket code from invoke_graph_rag.py. Add document embeddings core lifecycle to the knowledge service: list/get/put/delete/load operations across schema, translator, Cassandra table store, knowledge manager, gateway registry, REST API, socket client, and CLI (tg-get-de-core, tg-put-de-core). Fix delete_kg_core to also clean up document embeddings rows. * Remove spurious workspace parameter from SPARQL algebra evaluator (#915) Fix threading of workspace paramater: - The SPARQL algebra evaluator was threading a workspace parameter through every function and passing it to TriplesClient.query(), which doesn't accept it. Workspace isolation is handled by pub/sub topic routing — the TriplesClient is already scoped to a workspace-specific flow, same as GraphRAG. Passing workspace explicitly was both incorrect and unnecessary. Update tests: - tests/unit/test_query/test_sparql_algebra.py (new) — Tests _query_pattern, _eval_bgp, and evaluate() with various algebra nodes. Key tests assert workspace is never in tc.query() kwargs, plus correctness tests for BGP, JOIN, UNION, SLICE, DISTINCT, and edge cases. - tests/unit/test_retrieval/test_graph_rag.py — Added test_triples_query_never_passes_workspace (checks query()) and test_follow_edges_never_passes_workspace (checks query_stream()). * Make all Cassandra and Qdrant I/O async-safe with proper concurrency controls (#916) 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 * Fixed error only returning a page of results (#921) The root cause: async_execute only materialises the first result page (by design — it says so in its docstring). The streaming query set fetch_size=20 and expected to iterate all results, but only got the first 20 rows back. The fix uses asyncio.to_thread(lambda: list(tg.session.execute(...))) which lets the sync driver iterate all pages in a worker thread — exactly what the pre-async code did. * Optional test warning suppression (#923) * Fix test collection module errors & silence upstream Pytest warnings (#823) * chore: add virtual environment and .env directories to gitignore * test: filter upstream DeprecationWarning and UserWarning messages * fix(namespace): remove empty __init__.py files to fix PEP 420 implicit namespace routing for trustgraph sub-packages * Revert __init__.py deletions * Add .ini changes but commented out, will be useful at times --------- Co-authored-by: Salil M <d2kyt@protonmail.com>
2026-05-15 13:02:51 +01:00
proc._cache_lock = asyncio.Lock()
proc._known_collections = set()
msg = MagicMock()
msg.metadata.collection = "col1"
entity = MagicMock()
entity.entity = Term(type=IRI, iri="http://example.org/Alice")
entity.vector = [0.1, 0.2, 0.3]
entity.chunk_id = "c1"
msg.entities = [entity]
2026-04-22 15:19:57 +01:00
await proc.store_graph_embeddings("user1", msg)
proc.qdrant.upsert.assert_called_once()
@pytest.mark.asyncio
async def test_lazy_collection_creation_on_new_dimension(self):
release/v2.4 -> master (#924) * CLI auth migration, document embeddings core lifecycle (#913) Migrate get_kg_core and put_kg_core CLI tools to use Api/SocketClient with first-frame auth (fixes broken raw websocket path). Fix wire format field names (root/vector). Remove ~600 lines of dead raw websocket code from invoke_graph_rag.py. Add document embeddings core lifecycle to the knowledge service: list/get/put/delete/load operations across schema, translator, Cassandra table store, knowledge manager, gateway registry, REST API, socket client, and CLI (tg-get-de-core, tg-put-de-core). Fix delete_kg_core to also clean up document embeddings rows. * Remove spurious workspace parameter from SPARQL algebra evaluator (#915) Fix threading of workspace paramater: - The SPARQL algebra evaluator was threading a workspace parameter through every function and passing it to TriplesClient.query(), which doesn't accept it. Workspace isolation is handled by pub/sub topic routing — the TriplesClient is already scoped to a workspace-specific flow, same as GraphRAG. Passing workspace explicitly was both incorrect and unnecessary. Update tests: - tests/unit/test_query/test_sparql_algebra.py (new) — Tests _query_pattern, _eval_bgp, and evaluate() with various algebra nodes. Key tests assert workspace is never in tc.query() kwargs, plus correctness tests for BGP, JOIN, UNION, SLICE, DISTINCT, and edge cases. - tests/unit/test_retrieval/test_graph_rag.py — Added test_triples_query_never_passes_workspace (checks query()) and test_follow_edges_never_passes_workspace (checks query_stream()). * Make all Cassandra and Qdrant I/O async-safe with proper concurrency controls (#916) 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 * Fixed error only returning a page of results (#921) The root cause: async_execute only materialises the first result page (by design — it says so in its docstring). The streaming query set fetch_size=20 and expected to iterate all results, but only got the first 20 rows back. The fix uses asyncio.to_thread(lambda: list(tg.session.execute(...))) which lets the sync driver iterate all pages in a worker thread — exactly what the pre-async code did. * Optional test warning suppression (#923) * Fix test collection module errors & silence upstream Pytest warnings (#823) * chore: add virtual environment and .env directories to gitignore * test: filter upstream DeprecationWarning and UserWarning messages * fix(namespace): remove empty __init__.py files to fix PEP 420 implicit namespace routing for trustgraph sub-packages * Revert __init__.py deletions * Add .ini changes but commented out, will be useful at times --------- Co-authored-by: Salil M <d2kyt@protonmail.com>
2026-05-15 13:02:51 +01:00
import asyncio
from trustgraph.storage.graph_embeddings.qdrant.write import Processor
proc = Processor.__new__(Processor)
proc.qdrant = MagicMock()
proc.qdrant.collection_exists.return_value = False
proc.collection_exists = MagicMock(return_value=True)
release/v2.4 -> master (#924) * CLI auth migration, document embeddings core lifecycle (#913) Migrate get_kg_core and put_kg_core CLI tools to use Api/SocketClient with first-frame auth (fixes broken raw websocket path). Fix wire format field names (root/vector). Remove ~600 lines of dead raw websocket code from invoke_graph_rag.py. Add document embeddings core lifecycle to the knowledge service: list/get/put/delete/load operations across schema, translator, Cassandra table store, knowledge manager, gateway registry, REST API, socket client, and CLI (tg-get-de-core, tg-put-de-core). Fix delete_kg_core to also clean up document embeddings rows. * Remove spurious workspace parameter from SPARQL algebra evaluator (#915) Fix threading of workspace paramater: - The SPARQL algebra evaluator was threading a workspace parameter through every function and passing it to TriplesClient.query(), which doesn't accept it. Workspace isolation is handled by pub/sub topic routing — the TriplesClient is already scoped to a workspace-specific flow, same as GraphRAG. Passing workspace explicitly was both incorrect and unnecessary. Update tests: - tests/unit/test_query/test_sparql_algebra.py (new) — Tests _query_pattern, _eval_bgp, and evaluate() with various algebra nodes. Key tests assert workspace is never in tc.query() kwargs, plus correctness tests for BGP, JOIN, UNION, SLICE, DISTINCT, and edge cases. - tests/unit/test_retrieval/test_graph_rag.py — Added test_triples_query_never_passes_workspace (checks query()) and test_follow_edges_never_passes_workspace (checks query_stream()). * Make all Cassandra and Qdrant I/O async-safe with proper concurrency controls (#916) 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 * Fixed error only returning a page of results (#921) The root cause: async_execute only materialises the first result page (by design — it says so in its docstring). The streaming query set fetch_size=20 and expected to iterate all results, but only got the first 20 rows back. The fix uses asyncio.to_thread(lambda: list(tg.session.execute(...))) which lets the sync driver iterate all pages in a worker thread — exactly what the pre-async code did. * Optional test warning suppression (#923) * Fix test collection module errors & silence upstream Pytest warnings (#823) * chore: add virtual environment and .env directories to gitignore * test: filter upstream DeprecationWarning and UserWarning messages * fix(namespace): remove empty __init__.py files to fix PEP 420 implicit namespace routing for trustgraph sub-packages * Revert __init__.py deletions * Add .ini changes but commented out, will be useful at times --------- Co-authored-by: Salil M <d2kyt@protonmail.com>
2026-05-15 13:02:51 +01:00
proc._cache_lock = asyncio.Lock()
proc._known_collections = set()
msg = MagicMock()
msg.metadata.collection = "graphs"
entity = MagicMock()
entity.entity = Term(type=IRI, iri="http://example.org/x")
entity.vector = [0.0] * 768
entity.chunk_id = ""
msg.entities = [entity]
2026-04-22 15:19:57 +01:00
await proc.store_graph_embeddings("alice", msg)
# Collection should be created with correct dimension
proc.qdrant.create_collection.assert_called_once()
create_args = proc.qdrant.create_collection.call_args
assert create_args[1]["collection_name"] == "t_alice_graphs_768"
# ---------------------------------------------------------------------------
# Collection validation — deleted-while-in-flight protection
# ---------------------------------------------------------------------------
class TestCollectionValidation:
@pytest.mark.asyncio
async def test_doc_embeddings_dropped_for_deleted_collection(self):
from trustgraph.storage.doc_embeddings.qdrant.write import Processor
proc = Processor.__new__(Processor)
proc.qdrant = MagicMock()
proc.collection_exists = MagicMock(return_value=False)
msg = MagicMock()
msg.metadata.collection = "deleted-col"
msg.chunks = [MagicMock()]
2026-04-22 15:19:57 +01:00
await proc.store_document_embeddings("user1", msg)
proc.qdrant.upsert.assert_not_called()
@pytest.mark.asyncio
async def test_graph_embeddings_dropped_for_deleted_collection(self):
from trustgraph.storage.graph_embeddings.qdrant.write import Processor
proc = Processor.__new__(Processor)
proc.qdrant = MagicMock()
proc.collection_exists = MagicMock(return_value=False)
msg = MagicMock()
msg.metadata.collection = "deleted-col"
msg.entities = [MagicMock()]
2026-04-22 15:19:57 +01:00
await proc.store_graph_embeddings("user1", msg)
proc.qdrant.upsert.assert_not_called()