fix(kb): 完善知识库入库和重建索引

review_agent/knowledge_base.py

@@ -10,8 +10,8 @@ from django.core.files.uploadedfile import UploadedFile
 
 from review_agent.models import KnowledgeBaseDocument
 from review_agent.regulatory_review.services.rag_citation import RagIndexUnavailable, retrieve_citations
-from review_agent.regulatory_review.services.rag_embedding import DeterministicEmbeddingProvider
-from review_agent.regulatory_review.services.rag_index import chunk_text, extract_text_from_path
+from review_agent.regulatory_review.services.rag_embedding import get_embedding_provider
+from review_agent.regulatory_review.services.rag_index import chunk_text, extract_text_from_path, is_excluded_source_path
 from review_agent.regulatory_review.services.rule_loader import DEFAULT_RULE_PATH, compute_file_sha256, load_rule_file
 
 
@@ -78,6 +78,8 @@ def list_source_documents(source_dir: Path) -> list[dict[str, Any]]:
             continue
         suffix = path.suffix.lower()
         relative_path = str(path.relative_to(source_dir))
+        if is_excluded_source_path(relative_path):
+            continue
         indexed_chunk_count = source_chunk_counts.get(relative_path, 0)
         documents.append(
             {
@@ -101,7 +103,7 @@ def search_knowledge_base(query: str, *, n_results: int = 3) -> dict[str, Any]:
     try:
         results = retrieve_citations(
             normalized,
-            embedding_provider=DeterministicEmbeddingProvider(),
+            embedding_provider=get_embedding_provider(),
             n_results=n_results,
         )
     except RagIndexUnavailable as exc:
@@ -210,7 +212,7 @@ def index_managed_document(document: KnowledgeBaseDocument) -> int:
             return 0
         collection = _load_chroma_collection()
         texts = [chunk.text for chunk in chunks]
-        embeddings = DeterministicEmbeddingProvider()(texts)
+        embeddings = get_embedding_provider()(texts)
         ids = [
             hashlib.sha256(f"managed:{document.pk}:{chunk.metadata['chunk_index']}".encode("utf-8")).hexdigest()
             for chunk in chunks