增加代码知识库；修复文档处理内容；增加API设置

rag-web-ui/backend/app/services/code_kb/schema.py

@@ -0,0 +1,63 @@
+from __future__ import annotations
+
+from dataclasses import asdict, dataclass, field
+from typing import Any, Dict, List, Optional
+
+
+@dataclass
+class CodeFunctionEvidence:
+    node_id: str
+    name: str
+    qualified_name: str
+    file: str
+    start_line: Optional[int] = None
+    end_line: Optional[int] = None
+    signature: str = ""
+    summary: str = ""
+    logic_flow: str = ""
+    code_snippet: str = ""
+    calls: List[str] = field(default_factory=list)
+    called_by: List[str] = field(default_factory=list)
+    includes: List[str] = field(default_factory=list)
+    embedding_model: str = ""
+    embedding_dim: int = 0
+    embedding_available: bool = True
+    raw: Dict[str, Any] = field(default_factory=dict)
+
+    def to_dict(self) -> Dict[str, Any]:
+        return asdict(self)
+
+
+@dataclass
+class CodeSearchHit:
+    evidence: CodeFunctionEvidence
+    similarity: float
+    distance: float
+    rank: int
+
+    def to_dict(self) -> Dict[str, Any]:
+        data = self.evidence.to_dict()
+        data.update(
+            {
+                "similarity": self.similarity,
+                "distance": self.distance,
+                "rank": self.rank,
+            }
+        )
+        return data
+
+
+@dataclass
+class CodeGraphContext:
+    node_id: str
+    callers: List[CodeFunctionEvidence] = field(default_factory=list)
+    callees: List[CodeFunctionEvidence] = field(default_factory=list)
+    call_chains: List[str] = field(default_factory=list)
+
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "node_id": self.node_id,
+            "callers": [item.to_dict() for item in self.callers],
+            "callees": [item.to_dict() for item in self.callees],
+            "call_chains": self.call_chains,
+        }