1026b1d871
- 后端: milvus_graph_service 增加图检索能力; benchmark_generation 重构为基于 service 的模式 - 后端: evaluation_service.py 迁移至 knowledge/eval/service.py - 后端: 简化 milvus.py 实现, 抽离图逻辑到 graph service - 前端: QuerySection.vue 大幅重构, 支持图检索模式与多视图切换 - 前端: BenchmarkGenerateModal 增加图检索配置选项 - 测试: 新增 service_generation 测试, 更新 benchmark_generation 测试
52 lines
2.2 KiB
Python
52 lines
2.2 KiB
Python
from yuxi.knowledge.graphs.milvus_graph_service import MilvusGraphService
|
|
from yuxi.knowledge.implementations.milvus import MilvusKB, _retrieval_config_options
|
|
|
|
|
|
def test_milvus_retrieval_config_exposes_graph_and_dependencies():
|
|
options = _retrieval_config_options()
|
|
by_key = {option["key"]: option for option in options}
|
|
|
|
assert by_key["use_graph_retrieval"]["default"] is False
|
|
assert by_key["graph_max_nodes"]["default"] == 10000
|
|
assert by_key["graph_max_nodes"]["depend_on"] == ("use_graph_retrieval", True)
|
|
assert by_key["graph_top_k"]["depend_on"] == ("use_graph_retrieval", True)
|
|
assert by_key["reranker_model"]["depend_on"] == ("use_reranker", True)
|
|
|
|
|
|
def test_graph_ppr_ranks_chunk_nodes_from_seed_entities():
|
|
subgraph = {
|
|
"nodes": [
|
|
{"id": "e1", "type": "Entity", "properties": {"entity_id": "seed"}},
|
|
{"id": "c1", "type": "Chunk", "properties": {"chunk_id": "chunk_a"}},
|
|
{"id": "e2", "type": "Entity", "properties": {"entity_id": "other"}},
|
|
{"id": "c2", "type": "Chunk", "properties": {"chunk_id": "chunk_b"}},
|
|
],
|
|
"edges": [
|
|
{"source_id": "e1", "target_id": "c1"},
|
|
{"source_id": "e1", "target_id": "e2"},
|
|
{"source_id": "e2", "target_id": "c2"},
|
|
],
|
|
}
|
|
|
|
ranked = MilvusGraphService.rank_chunks_by_ppr(subgraph, {"seed": 1.0}, top_k=2, damping=0.85)
|
|
|
|
assert [chunk_id for chunk_id, _ in ranked] == ["chunk_a", "chunk_b"]
|
|
|
|
|
|
def test_rrf_fusion_merges_chunk_and_graph_rankings():
|
|
kb = object.__new__(MilvusKB)
|
|
base_chunks = [
|
|
{"content": "base a", "metadata": {"chunk_id": "a"}, "score": 0.9},
|
|
{"content": "base b", "metadata": {"chunk_id": "b"}, "score": 0.8},
|
|
]
|
|
graph_chunks = [
|
|
{"content": "graph b", "metadata": {"chunk_id": "b"}, "score": 0.7, "graph_score": 0.7},
|
|
{"content": "graph c", "metadata": {"chunk_id": "c"}, "score": 0.6, "graph_score": 0.6},
|
|
]
|
|
|
|
fused = kb._fuse_chunk_rankings(base_chunks, graph_chunks, graph_weight=1.0)
|
|
|
|
assert [chunk["metadata"]["chunk_id"] for chunk in fused] == ["b", "a", "c"]
|
|
assert fused[0]["graph_score"] == 0.7
|
|
assert fused[0]["fusion_sources"] == ["chunk", "graph"]
|