fix: restore embedded wiki images and optimize vector sync

offline-backend-20260512/backend/tests/test_llmwiki_local_index_repository.py

@@ -40,20 +40,31 @@ async def test_sql_store_atomically_replaces_vectors_and_marks_missing_deleted(t
         )
         await session.commit()
 
-    store = SqlLlmWikiIndexStore(session_factory)
+    store = SqlLlmWikiIndexStore(session_factory, write_batch_size=1)
     assert await store.try_acquire_sync_lease("mapping-1", owner="worker", sync_id="sync-1", lease_seconds=60, fingerprint="fp")
     assert not await store.try_acquire_sync_lease("mapping-1", owner="worker", sync_id="sync-2", lease_seconds=60, fingerprint="fp")
     page = normalize_wiki_page("mapping-1", {"id": "p1", "slug": "people/alice", "title": "Alice", "content": "Alice founded Acme.", "status": "published"})
     blob, dimensions = encode_vector([3.0, 4.0], dimensions=2)
     saved = await store.replace_page_vectors(
         page=page,
-        vectors=[{"section_index": 0, "heading": "History", "section_content": "Alice founded Acme.", "content_hash": "section", "embedding_dimensions": dimensions, "vector_blob": blob}],
+        vectors=[
+            {
+                "section_index": index,
+                "heading": "History",
+                "section_content": f"Alice founded Acme. Part {index}",
+                "content_hash": f"section-{index}",
+                "embedding_dimensions": dimensions,
+                "vector_blob": blob,
+            }
+            for index in range(3)
+        ],
         fingerprint="fp",
         sync_id="sync-1",
     )
     snapshot = await store.load_vector_snapshot("mapping-1", "fp")
     assert saved["slug"] == "people/alice"
-    assert len(snapshot["rows"]) == 1
+    assert store.write_batch_size == 1
+    assert len(snapshot["rows"]) == 3
     assert decode_vector(snapshot["rows"][0]["vector_blob"], 2).tolist() == pytest.approx([0.6, 0.8])
 
     assert await store.mark_missing_pages_deleted("mapping-1", "sync-2") == 1

offline-backend-20260512/backend/tests/test_llmwiki_weknora.py

@@ -690,6 +690,24 @@ async def test_weknora_client_proxies_original_document_preview_without_credenti
     assert media_type == "text/plain"
 
 
+@pytest.mark.asyncio
+async def test_weknora_client_fetches_protected_wiki_image() -> None:
+    async def handler(request: httpx.Request) -> httpx.Response:
+        assert request.url.path == "/api/v1/knowledge-bases/kb-one/files"
+        assert request.url.params["file_path"] == "resource://wiki/figure.png"
+        return httpx.Response(200, content=b"png-bytes", headers={"Content-Type": "image/png"})
+
+    async with httpx.AsyncClient(transport=httpx.MockTransport(handler)) as http_client:
+        client = WeKnoraClient(base_url="http://weknora.local:8080", http_client=http_client)
+        content, media_type = await client.get_knowledge_base_file(
+            "kb-one",
+            "resource://wiki/figure.png",
+        )
+
+    assert content == b"png-bytes"
+    assert media_type == "image/png"
+
+
 @pytest.mark.asyncio
 async def test_memory_store_enforces_owner_and_public_visibility() -> None:
     store = MemoryLlmWikiStore()
@@ -1358,6 +1376,10 @@ def test_weknora_iframe_html_rewrites_public_prefix_paths() -> None:
     assert '"allowedWeKnoraIds":["remote-kb","other-visible-kb"]' in injected
     assert ".breadcrumb-link.dropdown" not in injected
     assert "reloadForKnowledgeBaseSwitch" in injected
+    assert "function protectedImageApiUrl" in injected
+    assert 'img[data-protected-src]' in injected
+    assert 'encodeURIComponent(cfg.weknoraId)' in injected
+    assert 'credentials: "same-origin"' in injected
 
 
 def test_weknora_iframe_filters_kb_list_to_signed_visible_mappings() -> None:

offline-backend-20260512/backend/tests/test_llmwiki_wiki_sync.py

@@ -1,10 +1,13 @@
 from __future__ import annotations
 
+import asyncio
 import json
+from types import SimpleNamespace
 
 import httpx
 import pytest
 
+from app.gateway import llmwiki_index_scheduler
 from deerflow.config.llmwiki_config import LocalWikiEmbeddingConfig, LocalWikiIndexConfig
 from deerflow.integrations.weknora.local_index.embedding import StrictWikiEmbeddingClient
 from deerflow.integrations.weknora.local_index.sync import WikiSyncService
@@ -85,6 +88,24 @@ async def test_configured_dimensions_are_validated_but_not_sent_to_bge() -> None
     assert requests == [{"model": "bge-embedding-m3", "input": ["Wiki page"]}]
 
 
+@pytest.mark.asyncio
+async def test_embedding_error_includes_upstream_reason(caplog) -> None:
+    def handler(request: httpx.Request) -> httpx.Response:
+        return httpx.Response(400, json={"detail": "unsupported request parameter: dimensions"})
+
+    config = LocalWikiEmbeddingConfig(
+        model="bge-embedding-m3",
+        base_url="http://embedding.test/v1",
+        max_retries=0,
+    )
+    async with httpx.AsyncClient(transport=httpx.MockTransport(handler)) as http_client:
+        client = StrictWikiEmbeddingClient(config, http_client=http_client)
+        with pytest.raises(RuntimeError, match="unsupported request parameter: dimensions"):
+            await client.embed_query("Wiki page")
+
+    assert "unsupported request parameter: dimensions" in caplog.text
+
+
 @pytest.mark.asyncio
 async def test_omitted_dimensions_are_discovered_from_first_response() -> None:
     def handler(request: httpx.Request) -> httpx.Response:
@@ -167,3 +188,39 @@ async def test_fingerprint_change_automatically_rebuilds_entire_mapping() -> Non
     assert store.states["kb"]["state"] == "idle"
     assert store.states["kb"]["embedding_fingerprint"] == "fp-new"
     assert {vector["embedding_fingerprint"] for vectors in store.vectors.values() for vector in vectors} == {"fp-new"}
+
+
+@pytest.mark.asyncio
+async def test_auto_sync_scheduler_waits_before_first_scan(monkeypatch) -> None:
+    calls: list[str] = []
+
+    async def fake_sleep(seconds: float) -> None:
+        calls.append(f"sleep:{seconds}")
+        raise asyncio.CancelledError
+
+    async def fake_sync_once(app) -> None:
+        calls.append("sync")
+
+    monkeypatch.setattr(llmwiki_index_scheduler.asyncio, "sleep", fake_sleep)
+    monkeypatch.setattr(llmwiki_index_scheduler, "_sync_once", fake_sync_once)
+    app = SimpleNamespace(
+        state=SimpleNamespace(
+            config=SimpleNamespace(
+                llmwiki=SimpleNamespace(
+                    local_wiki_index=SimpleNamespace(sync_interval_seconds=120),
+                )
+            )
+        )
+    )
+
+    with pytest.raises(asyncio.CancelledError):
+        await llmwiki_index_scheduler.local_wiki_index_scheduler_loop(app)
+
+    assert calls == ["sleep:120"]
+
+
+def test_local_wiki_index_accepts_legacy_sync_flag() -> None:
+    config = LocalWikiIndexConfig.model_validate({"sync_enabled": True})
+
+    assert config.auto_sync is True
+    assert config.sync_enabled is True

offline-backend-20260512/backend/tests/test_public_llmwiki_embed.py

@@ -177,6 +177,37 @@ async def test_public_wiki_page_list_is_normalized(monkeypatch) -> None:
     assert payload["pages"][0]["aliases"] == []
 
 
+@pytest.mark.asyncio
+async def test_public_wiki_file_is_authorized_and_streamed(monkeypatch) -> None:
+    store = PublicStore(_published_row())
+
+    class Client:
+        async def get_knowledge_base_file(self, knowledge_base_id: str, file_path: str):
+            assert knowledge_base_id == "remote-one"
+            assert file_path == "storage://tenant/resource://wiki/figure.png"
+            return b"png-bytes", "image/png"
+
+    monkeypatch.setattr(llmwiki, "_client_or_503", lambda: Client())
+
+    response = await llmwiki.get_public_wiki_file(
+        _request(store),
+        "kb-one",
+        "storage://tenant/resource://wiki/figure.png",
+    )
+
+    assert store.calls == [("kb-one", "__public_wiki__", False, False)]
+    assert response.body == b"png-bytes"
+    assert response.media_type == "image/png"
+    assert response.headers["cache-control"] == "public, max-age=300"
+
+
+def test_wiki_file_proxy_rejects_non_weknora_urls() -> None:
+    with pytest.raises(HTTPException) as exc_info:
+        llmwiki._validate_protected_wiki_file_path("https://example.test/private.png")
+
+    assert exc_info.value.status_code == 422
+
+
 @pytest.mark.asyncio
 async def test_public_wiki_index_preserves_weknora_group_and_item_order(monkeypatch) -> None:
     store = PublicStore(_published_row())