fix(scanner): compare model roots by identity, not by spelling (#1146)

ComfyUI hands its model folders over in the host's own spelling --
`utils/extra_config.py` runs every configured path through `os.path.normpath()`,
which is all backslashes on Windows, and `main.py` registers `<output>/loras`
unconditionally for its save nodes. `Config` kept that spelling verbatim while
`_prepare_*_paths()` normalized the live root lists to forward slashes, and the
scanner compared the two as raw strings. On Windows every configured root
therefore looked like a root that was not in the live list: the Refresh menu
listed each directory twice (the twin showing "0 models"), the completion toast
claimed reachable roots were "currently inaccessible" and kept their whole cache
count as `kept_unreachable`, and the settings sync persisted host-registered
directories that were never created.

The duplication was not only cosmetic. `_UnreachablePaths.add()` normalizes the
prefix it is given, so the configured twin of a live root registered exactly the
live root's prefix as unreadable: in the prune phase every cached-but-missing
entry under it matched and was "kept", so models deleted from disk were never
removed from the cache again.

* `py/utils/root_paths.py` holds the single `normalize_root_identity()`
  (separator style, redundant/trailing separators, and case where the filesystem
  folds it: whole paths on Windows, drive letters everywhere). `config.py` and
  `settings_manager.py` each carried a private copy and `model_scanner.py`
  compared raw strings.
* `ModelScanner` compares root identities in `get_configured_model_roots()`,
  `describe_model_roots()` (dedupe + `available`), the reconcile's
  `configured_only` list and the `ReconcileScope.roots` membership test, so one
  directory is one root whatever the host called it. `_normalized_root_prefix()`
  is built on the same identity, and an empty prefix no longer matches every
  path.
* A configured root that never made it into the live list is reported only while
  it still holds cached entries -- in the menu and in `skipped_roots` /
  `unavailable_paths` alike. A directory that neither exists nor was ever scanned
  has nothing to keep, so `<output>/loras` stops being an unavailable root.
* The startup sync drops saved roots that neither exist nor hold cached entries
  (`PersistentModelCache.roots_with_cached_entries()`, one scan per model type,
  separator-agnostic and wildcard-escaped). A root named by the library's
  `default_*_root` is kept even when empty: dropping it would let the next boot
  "repair" the default onto another root. When the cache cannot be consulted,
  every root is kept.
* `GET /scan?roots=` validates requested roots by identity too, so the browser
  extension or a script passing the spelling settings.json holds selects the root
  instead of answering 400 "Unknown model root(s)".

Verified: 8 of the new tests fail against the pre-fix source (checked with the
fix stashed); 3798 passed, 7 skipped, 11 deselected. The probe costs 71ms for a
100k-model library with seven missing roots (357ms when probing root by root).
This commit is contained in:
Will Miao
2026-10-12 08:43:06 +08:00
parent e056142669
commit 6d0d7cb15b
11 changed files with 822 additions and 57 deletions
+130
View File
@@ -2949,3 +2949,133 @@ def test_reconcile_walk_skips_the_recipes_dir(tmp_path: Path, monkeypatch):
assert "keep" in result.discovered_folders
assert "recipes" not in result.discovered_folders
assert "recipes/sub" not in result.discovered_folders
# --- the host's path spelling (issue #1146) ---------------------------------
def _windows_spelling(path: Path) -> str:
"""Return a path the way ComfyUI hands it over on Windows.
``utils/extra_config.py`` normalizes every configured model folder with
``os.path.normpath``, so on Windows the configured list is all backslashes
while the live root list is forward-slashed.
"""
return str(path).replace("/", "\\")
async def _windows_spelling_scanner(tmp_path: Path, monkeypatch):
"""Scanner configured with backslash roots plus a never-created output dir.
``<output>/loras`` is what ComfyUI registers for its save nodes whether or not
the directory exists (``main.py`` calls ``add_model_folder_path`` with it
unconditionally).
"""
primary = tmp_path / "primary"
(primary / "pack").mkdir(parents=True)
(primary / "pack" / "a.txt").write_text("a", encoding="utf-8")
never_created = tmp_path / "output" / "loras"
scanner = MultiRootDummyScanner([primary])
await scanner._initialize_cache()
monkeypatch.setattr(
scanner,
"get_configured_model_roots",
lambda: [_windows_spelling(primary), _windows_spelling(never_created)],
)
return scanner, primary, never_created
@pytest.mark.asyncio
async def test_describe_roots_folds_the_host_path_spelling(tmp_path: Path, monkeypatch):
"""One directory is one root, whatever separator the host used."""
scanner, primary, _never_created = await _windows_spelling_scanner(
tmp_path, monkeypatch
)
described = scanner.describe_model_roots()
assert [entry["path"] for entry in described] == [_normalize_path(primary)]
assert described[0]["available"] is True
assert described[0]["reachable"] is True
assert described[0]["models"] == 1
@pytest.mark.asyncio
async def test_reconcile_does_not_shadow_a_backslash_configured_root(
tmp_path: Path, monkeypatch
):
"""A configured twin must not register its live root as unreachable.
``_UnreachablePaths`` normalizes the prefixes it is given, so the backslash
spelling of a root the walk just read normalized to the same prefix: every
cached entry under it counted as "kept", which is what made the refresh toast
claim thousands of reachable models were inaccessible -- and what stopped
deleted models from ever being pruned.
"""
scanner, primary, _never_created = await _windows_spelling_scanner(
tmp_path, monkeypatch
)
(primary / "pack" / "a.txt").unlink()
summary = await scanner._reconcile_cache()
assert summary is not None
assert summary["skipped_roots"] == []
assert summary["unavailable_paths"] == []
assert summary["kept_unreachable"] == 0
assert summary["removed"] == 1
assert scanner._cache.raw_data == []
@pytest.mark.asyncio
async def test_root_that_never_existed_is_not_reported(tmp_path: Path, monkeypatch):
"""A registered directory that neither exists nor holds cache is not a root.
Only roots the user really configured and whose models are cached are worth an
offline row or a "models kept" notice.
"""
primary = tmp_path / "primary"
(primary / "pack").mkdir(parents=True)
(primary / "pack" / "a.txt").write_text("a", encoding="utf-8")
never_created = tmp_path / "output" / "loras"
scanner = MultiRootDummyScanner([primary])
await scanner._initialize_cache()
monkeypatch.setattr(
scanner,
"get_configured_model_roots",
lambda: [str(primary), str(never_created)],
)
assert [entry["path"] for entry in scanner.describe_model_roots()] == [
_normalize_path(primary)
]
summary = await scanner._reconcile_cache()
assert summary is not None
assert summary["skipped_roots"] == []
assert summary["kept_unreachable"] == 0
@pytest.mark.asyncio
async def test_scoped_scan_accepts_the_host_path_spelling(tmp_path: Path, monkeypatch):
"""A caller may name a root the way the host spells it (issue #1146).
The Refresh menu sends back the path `/roots` rendered, but the browser
extension and scripts can pass what settings.json holds -- backslashes on
Windows -- and that must select the root rather than scan nothing.
"""
primary = tmp_path / "primary"
(primary / "pack").mkdir(parents=True)
(primary / "pack" / "a.txt").write_text("a", encoding="utf-8")
scanner = MultiRootDummyScanner([primary])
await scanner._initialize_cache()
scope = ReconcileScope(roots=(_windows_spelling(primary),))
summary = await scanner._reconcile_cache(scope=scope)
assert summary is not None
assert summary["scanned_roots"] == [primary.name]