perf(recipe): precompute image_id_map for O(1) CivitAI image existence checks

Build a civitai_image_id → recipe_id mapping once during cache
initialization instead of scanning all recipes on every
check_image_exists and import_from_url call.

- RecipeCache gains an image_id_map field populated by
  _build_image_id_map() during cache init
- check_image_exists and import_from_url duplicate detection
  now use the precomputed map (O(k) / O(1) vs O(n))
- Map is persisted in SQLite cache_metadata for fast startup
- Incrementally updated on add/remove/bulk_remove paths
- Fix: conn.close() before cache_metadata query (dead connection)

tests/routes/test_recipe_routes.py

@@ -46,6 +46,7 @@ class StubRecipeScanner:
         self.last_paginated_params: Dict[str, Any] | None = None
         self.lora_lookup: Dict[str, List[Dict[str, Any]]] = {}
         self.checkpoint_lookup: Dict[str, List[Dict[str, Any]]] = {}
+        self.image_id_map_override: Dict[str, str] = {}
 
         async def _noop_get_cached_data(force_refresh: bool = False) -> None:  # noqa: ARG001 - signature mirrors real scanner
             return None
@@ -56,7 +57,10 @@ class StubRecipeScanner:
         )
 
     async def get_cached_data(self, force_refresh: bool = False) -> SimpleNamespace:  # noqa: ARG002 - flag unused by stub
-        return SimpleNamespace(raw_data=list(self.cached_raw))
+        return SimpleNamespace(
+            raw_data=list(self.cached_raw),
+            image_id_map=dict(getattr(self, "image_id_map_override", {})),
+        )
 
     async def get_paginated_data(self, **params: Any) -> Dict[str, Any]:
         self.last_paginated_params = params
@@ -999,3 +1003,95 @@ async def test_batch_import_cancel_missing_id(monkeypatch, tmp_path: Path) -> No
         payload = await response.json()
         assert response.status == 400
         assert payload["success"] is False
+
+
+async def test_check_image_exists_uses_image_id_map(monkeypatch, tmp_path: Path) -> None:
+    """check_image_exists must use precomputed image_id_map instead of scanning raw_data."""
+    async with recipe_harness(monkeypatch, tmp_path) as harness:
+        harness.scanner.image_id_map_override = {
+            "123": "recipe-alpha",
+            "789": "recipe-gamma",
+        }
+
+        response = await harness.client.get(
+            "/api/lm/recipes/check-image-exists",
+            params={"image_ids": "123,456,789"},
+        )
+        payload = await response.json()
+
+        assert response.status == 200
+        assert payload["success"] is True
+        assert payload["results"]["123"] == {
+            "in_library": True,
+            "recipe_id": "recipe-alpha",
+        }
+        assert payload["results"]["456"] == {
+            "in_library": False,
+            "recipe_id": None,
+        }
+        assert payload["results"]["789"] == {
+            "in_library": True,
+            "recipe_id": "recipe-gamma",
+        }
+
+
+async def test_check_image_exists_handles_empty_input(monkeypatch, tmp_path: Path) -> None:
+    """Empty or non-numeric image_ids must return an empty results dict."""
+    async with recipe_harness(monkeypatch, tmp_path) as harness:
+        response = await harness.client.get(
+            "/api/lm/recipes/check-image-exists",
+            params={"image_ids": ""},
+        )
+        payload = await response.json()
+        assert response.status == 200
+        assert payload["results"] == {}
+
+
+async def test_import_from_url_detects_duplicate_via_image_id_map(
+    monkeypatch, tmp_path: Path,
+) -> None:
+    """import_from_url must return already_exists when image_id is in image_id_map."""
+    async with recipe_harness(monkeypatch, tmp_path) as harness:
+        harness.scanner.cached_raw = [
+            {"id": "existing-recipe", "title": "My Recipe"},
+        ]
+        harness.scanner.image_id_map_override = {
+            "99999": "existing-recipe",
+        }
+
+        response = await harness.client.get(
+            "/api/lm/recipes/import-from-url",
+            params={"image_url": "https://civitai.com/images/99999"},
+        )
+        payload = await response.json()
+
+        assert response.status == 200
+        assert payload["already_exists"] is True
+        assert payload["recipe_id"] == "existing-recipe"
+        assert payload["name"] == "My Recipe"
+
+
+async def test_import_from_url_proceeds_when_image_id_not_in_map(
+    monkeypatch, tmp_path: Path,
+) -> None:
+    """When image_id is absent from image_id_map, import_from_url must proceed to import."""
+    async with recipe_harness(monkeypatch, tmp_path) as harness:
+        harness.scanner.image_id_map_override = {
+            "111": "some-other-recipe",
+        }
+        harness.civitai.image_info["99999"] = {
+            "id": 99999,
+            "url": "https://image.civitai.com/x/y/original=true/sample.jpeg",
+            "type": "image",
+            "meta": {"prompt": "test"},
+        }
+
+        response = await harness.client.get(
+            "/api/lm/recipes/import-from-url",
+            params={"image_url": "https://civitai.com/images/99999"},
+        )
+
+        # The import may succeed or fail depending on downstream stubs,
+        # but it must NOT return already_exists
+        payload = await response.json()
+        assert payload.get("already_exists") is not True

tests/services/test_recipe_scanner.py

@@ -1015,3 +1015,85 @@ async def test_get_paginated_data_sorting(recipe_scanner):
     # Test Date ASC: Gamma (5), Alpha (10), Beta (20)
     res = await scanner.get_paginated_data(page=1, page_size=10, sort_by="date:asc")
     assert [i["id"] for i in res["items"]] == ["C", "A", "B"]
+
+
+async def test_build_image_id_map_filters_correctly(recipe_scanner):
+    """Only recipes with valid CivitAI source_path appear in image_id_map.
+
+    Recipes imported from local files or with empty/missing source_path
+    must be naturally excluded.
+    """
+    scanner, _ = recipe_scanner
+    from py.services.recipe_cache import RecipeCache
+
+    scanner._cache = RecipeCache(
+        raw_data=[
+            {"id": "r1", "source_path": "https://civitai.com/images/12345"},
+            {"id": "r2", "source_path": "https://civitai.com/images/67890"},
+            {"id": "r3", "source_path": "/home/user/local_image.png"},
+            {"id": "r4", "source_path": ""},
+            {"id": "r5"},
+        ],
+        sorted_by_name=[],
+        sorted_by_date=[],
+    )
+
+    result = scanner._build_image_id_map()
+
+    assert result == {
+        "12345": "r1",
+        "67890": "r2",
+    }
+    # r3 = local file path, r4 = empty string, r5 = no key → all excluded
+    for rid in ("r3", "r4", "r5"):
+        assert rid not in result.values()
+
+
+async def test_add_recipe_updates_image_id_map(recipe_scanner):
+    """Adding a recipe with a CivitAI URL must update image_id_map.
+
+    A recipe with a local file path must NOT produce an entry.
+    """
+    scanner, _ = recipe_scanner
+
+    await scanner.add_recipe({
+        "id": "civitai-recipe",
+        "title": "CivitAI",
+        "source_path": "https://civitai.com/images/55555",
+    })
+
+    cache = await scanner.get_cached_data()
+    assert cache.image_id_map.get("55555") == "civitai-recipe"
+
+    await scanner.add_recipe({
+        "id": "local-recipe",
+        "title": "Local",
+        "source_path": "/path/to/local.png",
+    })
+
+    assert "local-recipe" not in cache.image_id_map.values()
+
+
+async def test_remove_recipe_clears_image_id_map(recipe_scanner):
+    """Removing a recipe that has a CivitAI image_id must clean up the map."""
+    scanner, _ = recipe_scanner
+
+    await scanner.add_recipe({
+        "id": "recipe-a",
+        "title": "A",
+        "source_path": "https://civitai.com/images/111",
+    })
+    await scanner.add_recipe({
+        "id": "recipe-b",
+        "title": "B",
+        "source_path": "https://civitai.com/images/222",
+    })
+
+    cache = await scanner.get_cached_data()
+    assert "111" in cache.image_id_map
+    assert cache.image_id_map["222"] == "recipe-b"
+
+    await scanner.remove_recipe("recipe-a")
+
+    assert "111" not in cache.image_id_map
+    assert cache.image_id_map["222"] == "recipe-b"

tests/test_persistent_recipe_cache.py

@@ -465,3 +465,81 @@ class TestPersistentRecipeCache:
         # Operations should complete
         assert operation_counts["saves"] == 5
         assert operation_counts["removes"] == 5
+
+    # -----------------------------------------------------------------------
+    # image_id_map persistence (Phase 1 improvement)
+    # -----------------------------------------------------------------------
+
+    def test_save_and_load_image_id_map_roundtrip(self, temp_db_path, sample_recipes):
+        """Save image_id_map via save_cache() and verify it round-trips through load_cache()."""
+        cache = PersistentRecipeCache(db_path=temp_db_path)
+
+        image_id_map = {
+            "12345": "recipe-alpha",
+            "67890": "recipe-beta",
+        }
+        cache.save_cache(sample_recipes, image_id_map=image_id_map)
+
+        loaded = cache.load_cache()
+        assert loaded is not None
+        assert loaded.image_id_map == image_id_map
+
+    def test_load_without_image_id_map_returns_empty_dict(self, temp_db_path, sample_recipes):
+        """Loading from a cache that has no image_id_map metadata must yield {}."""
+        cache = PersistentRecipeCache(db_path=temp_db_path)
+
+        # Save without image_id_map
+        cache.save_cache(sample_recipes)
+
+        loaded = cache.load_cache()
+        assert loaded is not None
+        assert loaded.image_id_map == {}
+
+    def test_save_cache_without_image_id_map_does_not_corrupt_existing(
+        self, temp_db_path, sample_recipes,
+    ):
+        """Overwriting cache without passing image_id_map must not leave stale data.
+
+        The previous image_id_map entry in cache_metadata should be replaced with {}.
+        """
+        cache = PersistentRecipeCache(db_path=temp_db_path)
+
+        cache.save_cache(sample_recipes, image_id_map={"123": "old-recipe"})
+        # Overwrite without image_id_map
+        cache.save_cache(sample_recipes)
+
+        loaded = cache.load_cache()
+        assert loaded.image_id_map == {}
+
+    def test_image_id_map_survives_recipe_update(self, temp_db_path, sample_recipes):
+        """Updating a single recipe must not drop the image_id_map metadata."""
+        cache = PersistentRecipeCache(db_path=temp_db_path)
+
+        cache.save_cache(sample_recipes, image_id_map={"123": "recipe-alpha"})
+
+        updated = dict(sample_recipes[0])
+        updated["title"] = "Updated"
+        cache.update_recipe(updated)
+
+        loaded = cache.load_cache()
+        assert loaded.image_id_map == {"123": "recipe-alpha"}
+
+    def test_save_image_id_map_persists_without_full_save(self, temp_db_path, sample_recipes):
+        """save_image_id_map must update cache_metadata without rewriting all recipes."""
+        cache = PersistentRecipeCache(db_path=temp_db_path)
+        cache.save_cache(sample_recipes)
+
+        cache.save_image_id_map({"555": "new-recipe", "666": "another-recipe"})
+
+        loaded = cache.load_cache()
+        assert loaded.image_id_map == {"555": "new-recipe", "666": "another-recipe"}
+
+    def test_save_image_id_map_overwrites_previous(self, temp_db_path, sample_recipes):
+        """Calling save_image_id_map twice must replace, not merge."""
+        cache = PersistentRecipeCache(db_path=temp_db_path)
+        cache.save_cache(sample_recipes, image_id_map={"111": "old"})
+
+        cache.save_image_id_map({"222": "new-only"})
+
+        loaded = cache.load_cache()
+        assert loaded.image_id_map == {"222": "new-only"}