Backup automatico script del 2026-08-23 12:04

services/loogle-mcp/scripts/test_gitea_rag_p3.py

@@ -0,0 +1,130 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""Test P3: RAG semantico su file Gitea indicizzati."""
+
+from __future__ import annotations
+
+import json
+import os
+import sys
+
+ROOT = os.path.abspath(os.path.join(os.path.dirname(__file__), ".."))
+if ROOT not in sys.path:
+    sys.path.insert(0, ROOT)
+
+
+def _load_dotenv() -> None:
+    env_path = os.path.join(ROOT, ".env")
+    if not os.path.isfile(env_path):
+        return
+    with open(env_path, encoding="utf-8") as fh:
+        for line in fh:
+            line = line.strip()
+            if not line or line.startswith("#") or "=" not in line:
+                continue
+            key, _, value = line.partition("=")
+            key = key.strip()
+            if key and key not in os.environ:
+                os.environ[key] = value.strip().strip("'").strip('"')
+
+
+def _configure_paths() -> str:
+    in_container = os.path.isfile("/.dockerenv") or (
+        os.path.isdir("/data") and os.access("/data", os.W_OK)
+    )
+    if in_container:
+        os.environ.setdefault("MCP_DB", "/data/loogle_mcp.db")
+        os.environ.setdefault("MCP_VECTOR_FALLBACK", "/data/vector_fallback.db")
+        return "container"
+    os.environ["MCP_DB"] = "/tmp/loogle_mcp_p3_test.db"
+    os.environ["MCP_VECTOR_FALLBACK"] = "/tmp/loogle_mcp_p3_vectors.db"
+    return "host"
+
+
+def main() -> int:
+    _load_dotenv()
+    mode = _configure_paths()
+    print(f"mode={mode} db={os.environ['MCP_DB']} vector={os.environ['MCP_VECTOR_FALLBACK']}")
+
+    from app.db import get_conn, init_db
+    from app.knowledge import gitea_indexer
+    from app.mcp import tools
+
+    init_db()
+    if not get_conn().execute(
+        "SELECT 1 FROM sqlite_master WHERE name='indexed_gitea_files'"
+    ).fetchone():
+        raise RuntimeError("tabella indexed_gitea_files assente")
+
+    if not os.environ.get("GITEA_API_TOKEN_DANIELE", "").strip():
+        print("SKIP: GITEA_API_TOKEN_DANIELE assente")
+        return 1
+
+    repo = "daniele/rete"
+    existing = gitea_indexer.list_indexed_files(limit=5, repo=repo)
+    stats = gitea_indexer.index_stats()
+    if len(existing) < 1 or stats.get("qdrant_points", 0) < 1:
+        index_result = gitea_indexer.index_repo(
+            repo,
+            username="daniele",
+            force=True,
+            max_files=1,
+        )
+        print("index", json.dumps(index_result, ensure_ascii=False))
+    else:
+        print("index_skip", "already", len(existing), "files")
+
+    files = gitea_indexer.list_indexed_files(limit=5, repo=repo)
+    if not files:
+        raise RuntimeError("indexed_gitea_files vuota")
+    print("indexed_sample", [f["path"] for f in files[:3]])
+
+    claims = {"sub": "daniele", "scope": "knowledge:read gitea:read admin", "admin": "admin"}
+    for query in ("failover tier-b", "runbook"):
+        hits = gitea_indexer.search_gitea_knowledge("daniele", query, limit=5, is_admin=True)
+        if not hits:
+            raise RuntimeError(f"search_gitea_knowledge senza risultati per: {query}")
+        top = hits[0]
+        print(
+            f"search_ok[{query}]",
+            top.get("path") or top.get("title"),
+            round(float(top.get("score", 0)), 3),
+        )
+
+    tool_out = tools.call_tool(
+        "search_gitea_knowledge",
+        {"query": "failover", "limit": 5},
+        claims,
+    )
+    if not json.loads(tool_out["content"][0]["text"]).get("results"):
+        raise RuntimeError("tool search_gitea_knowledge vuoto")
+
+    unified = tools.call_tool(
+        "search_knowledge",
+        {"query": "FAILOVER CENSIMENTO documenti servizi", "limit": 12},
+        claims,
+    )
+    unified_payload = json.loads(unified["content"][0]["text"])
+    results = unified_payload.get("results") or []
+    sources = set()
+    for r in results:
+        if r.get("repo"):
+            sources.add("gitea")
+        elif r.get("source"):
+            sources.add(r["source"])
+        else:
+            sources.add("paperless")
+    print("search_knowledge_sources", sorted(sources))
+    if "gitea" not in sources:
+        raise RuntimeError("search_knowledge non include risultati Gitea")
+
+    list_tool = tools.call_tool("list_gitea_indexed_files", {"repo": repo, "limit": 5}, claims)
+    if json.loads(list_tool["content"][0]["text"]).get("count", 0) <= 0:
+        raise RuntimeError("list_gitea_indexed_files vuoto")
+
+    print("OK P3: search + unified + list (reindex opzionale via tool reindex_gitea_repo)")
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())