ComfyUI/tests-unit/assets_test/test_downloads.py
Simon Pinfold 55a15f87ce
feat(assets): add namespaced model_type tags and align tag semantics (#14511)
* feat(assets): add namespaced model type tags

* fix(assets): mark path-derived upload tags automatic

* fix(assets): merge duplicate scan specs

* test(assets): make duplicate path normalization portable

* feat(assets): add loader_path as the authoritative loader locator (#14796)

* fix(assets): filter model_type tags by bucket extension sets

Buckets sharing a base directory (e.g. diffusion_models and a custom
unet_gguf) tagged every file in the directory regardless of whether the
bucket could load it, so .safetensors files were tagged
model_type:unet_gguf and vice versa. Carry each bucket's registered
extension set through get_comfy_models_folders and only emit a
model_type tag when the file extension matches, keeping the empty-set
match-all convention from folder_paths.filter_files_extensions.

Files under a model base matching no bucket now keep only the models
tag instead of every directory-matching model_type tag.

* feat(assets): replace response file_path with persisted loader_path

The old file_path response field was a namespaced storage locator
(models/checkpoints/foo.safetensors): not an absolute path, not unique
identity, and not the value a loader consumes. Nothing needs that shape
on the wire (hash/ID-based locating is the long-term direction), so it
is dropped rather than renamed; the storage-root matching stays internal,
powering display_name.

What loaders DO need is the in-root loader path (category dropped:
models/checkpoints/foo/bar.safetensors -> foo/bar.safetensors). Serve it
as a first-class loader_path field, persisted on asset_references
(migration 0006) and written by every ingest pipeline at insert, so
responses read the column verbatim.

Like the model_type tags, loader_path is a seed-time derivative of the
model folder registry, maintained by the same scan lifecycle (new files seed
fresh values, pruning retires rows whose bucket disappeared). Rows
predating the column serve a null loader_path; databases from before
this stack already need recreating for the base branch's tag changes.

loader_path resolves every registered base including extra_model_paths
entries; display_name only the canonical storage roots. A file can
therefore be loadable with no display name (extra-path models) or the
reverse (unregistered files under the models root), and loader_path is
null exactly when no loader can resolve the file.

* test(assets): lock loader_path matrix (asymmetry, null, persist/read)

Cover the behaviour that has no production change but is easy to regress:
the extra-path asymmetry (loadable but no storage namespace), null
loader_path persistence for orphan files, and the response reading the
stored column with a compute fallback for un-backfilled rows.

* fix(assets): persist subfolder-qualified loader_path for ingested outputs

ingest_existing_file built its seed spec with the file's basename, so
outputs saved into a subfolder persisted loader_path (and the
user_metadata filename that preview URLs split for their subfolder
param) as just the basename: the served locator pointed at a file that
does not exist at that path. Scanner and seeder specs already derive
fname via compute_loader_path; use the same derivation here.

* fix(assets): only extension-matching buckets contribute a loader_path

The model-base match in get_asset_category_and_relative_path ignored
each bucket's extension set, so a file inside a registered base whose
extension the bucket cannot load (e.g. a .txt uploaded into
model_type:checkpoints) advertised a loader_path that no loader list
would ever resolve, while the tag side of the same stack already
excluded it. Apply the extension check used for backend tags (empty set
accepts any extension), keeping loader_path null exactly when no loader
can resolve the file.

* fix(assets): refresh loader_path when re-ingesting an existing reference

upsert_reference only wrote loader_path on the INSERT branch, so
re-ingesting an existing reference (an output overwritten in place, or a
file re-registered after its loader_path derivation changed) kept the
stale or NULL value forever. Write it on the UPDATE branch too, with a
null-safe change guard so a loader_path difference alone is enough to
trigger the update, and identical values stay a no-op.

* fix(assets): repair semantic merge breakage from #14796 and master

Two textually-clean but semantically-broken merges:

- routes.py lost its folder_paths import when #14796's import block
  superseded the base's, while the content-type hardening added via the
  base's master merge still calls folder_paths.is_dangerous_content_type.
- master's SVG download-hardening test uploads with the pre-namespacing
  bare checkpoints tag, which this branch's destination validation
  rejects; use model_type:checkpoints.

---------

Co-authored-by: guill <jacob.e.segal@gmail.com>
2026-07-08 22:00:08 -07:00

203 lines
7.5 KiB
Python

import contextlib
import json
import time
import uuid
from datetime import datetime
from pathlib import Path
from typing import Optional
import pytest
import requests
from helpers import get_asset_filename, trigger_sync_seed_assets
def test_download_svg_forced_to_attachment(http: requests.Session, api_base: str):
"""GHSA-779p-m5rp-r4h4 CISA-5 (sibling route): an uploaded SVG must never be
served inline from GET /api/assets/{id}/content, or an inline <script> runs
in the app origin (stored XSS). Even with disposition=inline requested, a
dangerous content type must be forced to application/octet-stream +
Content-Disposition: attachment + nosniff. Regression guard for the stale
inline blocklist that previously omitted image/svg+xml and ignored the
centralized folder_paths.is_dangerous_content_type check.
"""
svg = b'<svg xmlns="http://www.w3.org/2000/svg"><script>alert(1)</script></svg>'
files = {"file": ("evil.svg", svg, "image/svg+xml")}
form_data = {
"tags": json.dumps(["models", "model_type:checkpoints", "unit-tests", "svgxss"]),
"name": "evil.svg",
}
up = http.post(api_base + "/api/assets", files=files, data=form_data, timeout=120)
body = up.json()
assert up.status_code in (200, 201), body
aid = body["id"]
try:
r = http.get(f"{api_base}/api/assets/{aid}/content?disposition=inline", timeout=120)
r.content
assert r.status_code == 200
ct = r.headers.get("Content-Type", "").lower()
cd = r.headers.get("Content-Disposition", "").lower()
assert "svg" not in ct, f"SVG served with a renderable content type: {ct!r}"
assert ct.startswith("application/octet-stream"), f"expected octet-stream, got {ct!r}"
assert "attachment" in cd, f"inline disposition not overridden to attachment: {cd!r}"
assert r.headers.get("X-Content-Type-Options", "").lower() == "nosniff"
finally:
with contextlib.suppress(Exception):
http.delete(f"{api_base}/api/assets/{aid}", timeout=30)
def test_download_attachment_and_inline(http: requests.Session, api_base: str, seeded_asset: dict):
aid = seeded_asset["id"]
# default attachment
r1 = http.get(f"{api_base}/api/assets/{aid}/content", timeout=120)
data = r1.content
assert r1.status_code == 200
cd = r1.headers.get("Content-Disposition", "")
assert "attachment" in cd
assert data and len(data) == 4096
# inline requested
r2 = http.get(f"{api_base}/api/assets/{aid}/content?disposition=inline", timeout=120)
r2.content
assert r2.status_code == 200
cd2 = r2.headers.get("Content-Disposition", "")
assert "inline" in cd2
@pytest.mark.skip(reason="Requires computing hashes of files in directories to deduplicate into multiple cache states")
@pytest.mark.parametrize("root", ["input", "output"])
def test_download_chooses_existing_state_and_updates_access_time(
root: str,
http: requests.Session,
api_base: str,
comfy_tmp_base_dir: Path,
asset_factory,
make_asset_bytes,
run_scan_and_wait,
):
"""
Hashed asset with two state paths: if the first one disappears,
GET /content still serves from the remaining path and bumps last_access_time.
"""
scope = f"dl-first-{uuid.uuid4().hex[:6]}"
name = "first_existing_state.bin"
data = make_asset_bytes(name, 3072)
# Upload -> path1
a = asset_factory(name, [root, "unit-tests", scope], {}, data)
aid = a["id"]
base = comfy_tmp_base_dir / root / "unit-tests" / scope
path1 = base / get_asset_filename(a["asset_hash"], ".bin")
assert path1.exists()
# Seed path2 by copying, then scan to dedupe into a second state
path2 = base / "alt" / name
path2.parent.mkdir(parents=True, exist_ok=True)
path2.write_bytes(data)
trigger_sync_seed_assets(http, api_base)
run_scan_and_wait(root)
# Remove path1 so server must fall back to path2
path1.unlink()
# last_access_time before
rg0 = http.get(f"{api_base}/api/assets/{aid}", timeout=120)
d0 = rg0.json()
assert rg0.status_code == 200, d0
ts0 = d0.get("last_access_time")
time.sleep(0.05)
r = http.get(f"{api_base}/api/assets/{aid}/content", timeout=120)
blob = r.content
assert r.status_code == 200
assert blob == data # must serve from the surviving state (same bytes)
rg1 = http.get(f"{api_base}/api/assets/{aid}", timeout=120)
d1 = rg1.json()
assert rg1.status_code == 200, d1
ts1 = d1.get("last_access_time")
def _parse_iso8601(s: Optional[str]) -> Optional[float]:
if not s:
return None
s = s[:-1] if s.endswith("Z") else s
return datetime.fromisoformat(s).timestamp()
t0 = _parse_iso8601(ts0)
t1 = _parse_iso8601(ts1)
assert t1 is not None
if t0 is not None:
assert t1 > t0
@pytest.mark.parametrize("seeded_asset", [{"tags": ["models", "model_type:checkpoints"]}], indirect=True)
def test_download_missing_file_returns_404(
http: requests.Session, api_base: str, comfy_tmp_base_dir: Path, seeded_asset: dict
):
# Remove the underlying file then attempt download.
# We initialize fixture without additional tags to know exactly the asset file path.
try:
aid = seeded_asset["id"]
rg = http.get(f"{api_base}/api/assets/{aid}", timeout=120)
detail = rg.json()
assert rg.status_code == 200
asset_filename = get_asset_filename(detail["asset_hash"], ".safetensors")
abs_path = comfy_tmp_base_dir / "models" / "checkpoints" / asset_filename
assert abs_path.exists()
abs_path.unlink()
r2 = http.get(f"{api_base}/api/assets/{aid}/content", timeout=120)
assert r2.status_code == 404
body = r2.json()
assert body["error"]["code"] == "FILE_NOT_FOUND"
finally:
# We created asset without the "unit-tests" tag(see `autoclean_unit_test_assets`), we need to clear it manually.
dr = http.delete(f"{api_base}/api/assets/{aid}", timeout=120)
dr.content
@pytest.mark.skip(reason="Requires computing hashes of files in directories to deduplicate into multiple cache states")
@pytest.mark.parametrize("root", ["input", "output"])
def test_download_404_if_all_states_missing(
root: str,
http: requests.Session,
api_base: str,
comfy_tmp_base_dir: Path,
asset_factory,
make_asset_bytes,
run_scan_and_wait,
):
"""Multi-state asset: after the last remaining on-disk file is removed, download must return 404."""
scope = f"dl-404-{uuid.uuid4().hex[:6]}"
name = "missing_all_states.bin"
data = make_asset_bytes(name, 2048)
# Upload -> path1
a = asset_factory(name, [root, "unit-tests", scope], {}, data)
aid = a["id"]
base = comfy_tmp_base_dir / root / "unit-tests" / scope
p1 = base / get_asset_filename(a["asset_hash"], ".bin")
assert p1.exists()
# Seed a second state and dedupe
p2 = base / "copy" / name
p2.parent.mkdir(parents=True, exist_ok=True)
p2.write_bytes(data)
trigger_sync_seed_assets(http, api_base)
run_scan_and_wait(root)
# Remove first file -> download should still work via the second state
p1.unlink()
ok1 = http.get(f"{api_base}/api/assets/{aid}/content", timeout=120)
b1 = ok1.content
assert ok1.status_code == 200 and b1 == data
# Remove the last file -> download must 404
p2.unlink()
r2 = http.get(f"{api_base}/api/assets/{aid}/content", timeout=120)
body = r2.json()
assert r2.status_code == 404
assert body["error"]["code"] == "FILE_NOT_FOUND"