Add pruning of Assets not reachable through the current configs (#12168)

* Not sure about this one, but try removing assets from old sessions.

* Simplify _prune_orphaned_assets: merge functions, use list comprehensions

Amp-Thread-ID: https://ampcode.com/threads/T-019c0917-0dc3-75ab-870d-a32b3fdc1927
Co-authored-by: Amp <amp@ampcode.com>

* Refactor _prune_orphaned_assets for readability

Amp-Thread-ID: https://ampcode.com/threads/T-019c0917-0dc3-75ab-870d-a32b3fdc1927
Co-authored-by: Amp <amp@ampcode.com>

* Add unit tests for pruning

* Add unit tests for _prune_orphaned_assets

Tests cover:

- Orphaned seed assets pruned when file removed

- Seed assets with valid files survive

- Hashed assets not pruned even without file

- Multi-root pruning

- SQL LIKE escape handling for %, _, spaces

Amp-Thread-ID: https://ampcode.com/threads/T-019c0c7a-5c8a-7548-b6c3-823e9829ce74
Co-authored-by: Amp <amp@ampcode.com>

* Ruff fix

---------

Co-authored-by: Amp <amp@ampcode.com>

app/assets/scanner.py

@@ -27,6 +27,7 @@ def seed_assets(roots: tuple[RootType, ...], enable_logging: bool = False) -> No
     t_start = time.perf_counter()
     created = 0
     skipped_existing = 0
+    orphans_pruned = 0
     paths: list[str] = []
     try:
         existing_paths: set[str] = set()
@@ -38,6 +39,11 @@ def seed_assets(roots: tuple[RootType, ...], enable_logging: bool = False) -> No
             except Exception as e:
                 logging.exception("fast DB scan failed for %s: %s", r, e)
 
+        try:
+            orphans_pruned = _prune_orphaned_assets(roots)
+        except Exception as e:
+            logging.exception("orphan pruning failed: %s", e)
+
         if "models" in roots:
             paths.extend(collect_models_files())
         if "input" in roots:
@@ -85,15 +91,43 @@ def seed_assets(roots: tuple[RootType, ...], enable_logging: bool = False) -> No
     finally:
         if enable_logging:
             logging.info(
-                "Assets scan(roots=%s) completed in %.3fs (created=%d, skipped_existing=%d, total_seen=%d)",
+                "Assets scan(roots=%s) completed in %.3fs (created=%d, skipped_existing=%d, orphans_pruned=%d, total_seen=%d)",
                 roots,
                 time.perf_counter() - t_start,
                 created,
                 skipped_existing,
+                orphans_pruned,
                 len(paths),
             )
 
 
+def _prune_orphaned_assets(roots: tuple[RootType, ...]) -> int:
+    """Prune cache states outside configured prefixes, then delete orphaned seed assets."""
+    all_prefixes = [os.path.abspath(p) for r in roots for p in prefixes_for_root(r)]
+    if not all_prefixes:
+        return 0
+
+    def make_prefix_condition(prefix: str):
+        base = prefix if prefix.endswith(os.sep) else prefix + os.sep
+        escaped, esc = escape_like_prefix(base)
+        return AssetCacheState.file_path.like(escaped + "%", escape=esc)
+
+    matches_valid_prefix = sqlalchemy.or_(*[make_prefix_condition(p) for p in all_prefixes])
+
+    orphan_subq = (
+        sqlalchemy.select(Asset.id)
+        .outerjoin(AssetCacheState, AssetCacheState.asset_id == Asset.id)
+        .where(Asset.hash.is_(None), AssetCacheState.id.is_(None))
+    ).scalar_subquery()
+
+    with create_session() as sess:
+        sess.execute(sqlalchemy.delete(AssetCacheState).where(~matches_valid_prefix))
+        sess.execute(sqlalchemy.delete(AssetInfo).where(AssetInfo.asset_id.in_(orphan_subq)))
+        result = sess.execute(sqlalchemy.delete(Asset).where(Asset.id.in_(orphan_subq)))
+        sess.commit()
+        return result.rowcount
+
+
 def _fast_db_consistency_pass(
     root: RootType,
     *,

tests-assets/test_prune_orphaned_assets.py

@@ -0,0 +1,141 @@
+import uuid
+from pathlib import Path
+
+import pytest
+import requests
+from conftest import get_asset_filename, trigger_sync_seed_assets
+
+
+@pytest.fixture
+def create_seed_file(comfy_tmp_base_dir: Path):
+    """Create a file on disk that will become a seed asset after sync."""
+    created: list[Path] = []
+
+    def _create(root: str, scope: str, name: str | None = None, data: bytes = b"TEST") -> Path:
+        name = name or f"seed_{uuid.uuid4().hex[:8]}.bin"
+        path = comfy_tmp_base_dir / root / "unit-tests" / scope / name
+        path.parent.mkdir(parents=True, exist_ok=True)
+        path.write_bytes(data)
+        created.append(path)
+        return path
+
+    yield _create
+
+    for p in created:
+        p.unlink(missing_ok=True)
+
+
+@pytest.fixture
+def find_asset(http: requests.Session, api_base: str):
+    """Query API for assets matching scope and optional name."""
+    def _find(scope: str, name: str | None = None) -> list[dict]:
+        params = {"include_tags": f"unit-tests,{scope}"}
+        if name:
+            params["name_contains"] = name
+        r = http.get(f"{api_base}/api/assets", params=params, timeout=120)
+        assert r.status_code == 200
+        assets = r.json().get("assets", [])
+        if name:
+            return [a for a in assets if a.get("name") == name]
+        return assets
+
+    return _find
+
+
+@pytest.mark.parametrize("root", ["input", "output"])
+def test_orphaned_seed_asset_is_pruned(
+    root: str,
+    create_seed_file,
+    find_asset,
+    http: requests.Session,
+    api_base: str,
+):
+    """Seed asset with deleted file is removed; with file present, it survives."""
+    scope = f"prune-{uuid.uuid4().hex[:6]}"
+    fp = create_seed_file(root, scope)
+    name = fp.name
+
+    trigger_sync_seed_assets(http, api_base)
+    assert find_asset(scope, name), "Seed asset should exist"
+
+    fp.unlink()
+    trigger_sync_seed_assets(http, api_base)
+    assert not find_asset(scope, name), "Orphaned seed should be pruned"
+
+
+def test_seed_asset_with_file_survives_prune(
+    create_seed_file,
+    find_asset,
+    http: requests.Session,
+    api_base: str,
+):
+    """Seed asset with file still on disk is NOT pruned."""
+    scope = f"keep-{uuid.uuid4().hex[:6]}"
+    fp = create_seed_file("input", scope)
+
+    trigger_sync_seed_assets(http, api_base)
+    trigger_sync_seed_assets(http, api_base)
+
+    assert find_asset(scope, fp.name), "Seed with valid file should survive"
+
+
+def test_hashed_asset_not_pruned_when_file_missing(
+    http: requests.Session,
+    api_base: str,
+    comfy_tmp_base_dir: Path,
+    asset_factory,
+    make_asset_bytes,
+):
+    """Hashed assets are never deleted by prune, even without file."""
+    scope = f"hashed-{uuid.uuid4().hex[:6]}"
+    data = make_asset_bytes("test", 2048)
+    a = asset_factory("test.bin", ["input", "unit-tests", scope], {}, data)
+
+    path = comfy_tmp_base_dir / "input" / "unit-tests" / scope / get_asset_filename(a["asset_hash"], ".bin")
+    path.unlink()
+
+    trigger_sync_seed_assets(http, api_base)
+
+    r = http.get(f"{api_base}/api/assets/{a['id']}", timeout=120)
+    assert r.status_code == 200, "Hashed asset should NOT be pruned"
+
+
+def test_prune_across_multiple_roots(
+    create_seed_file,
+    find_asset,
+    http: requests.Session,
+    api_base: str,
+):
+    """Prune correctly handles assets across input and output roots."""
+    scope = f"multi-{uuid.uuid4().hex[:6]}"
+    input_fp = create_seed_file("input", scope, "input.bin")
+    create_seed_file("output", scope, "output.bin")
+
+    trigger_sync_seed_assets(http, api_base)
+    assert len(find_asset(scope)) == 2
+
+    input_fp.unlink()
+    trigger_sync_seed_assets(http, api_base)
+
+    remaining = find_asset(scope)
+    assert len(remaining) == 1
+    assert remaining[0]["name"] == "output.bin"
+
+
+@pytest.mark.parametrize("dirname", ["100%_done", "my_folder_name", "has spaces"])
+def test_special_chars_in_path_escaped_correctly(
+    dirname: str,
+    create_seed_file,
+    find_asset,
+    http: requests.Session,
+    api_base: str,
+    comfy_tmp_base_dir: Path,
+):
+    """SQL LIKE wildcards (%, _) and spaces in paths don't cause false matches."""
+    scope = f"special-{uuid.uuid4().hex[:6]}/{dirname}"
+    fp = create_seed_file("input", scope)
+
+    trigger_sync_seed_assets(http, api_base)
+    trigger_sync_seed_assets(http, api_base)
+
+    assert find_asset(scope.split("/")[0], fp.name), "Asset with special chars should survive"