feat(assets): add job_ids filter to GET /api/assets (#13998)

* feat(assets): add job_ids filter to GET /api/assets

Mirrors the existing cloud `job_ids` query param on the local Python server:
clients can pass a comma-separated list (or repeated query params) of UUIDs
to filter assets by their associated job.

The `AssetReference.job_id` column already exists, so no migration is
needed — this just plumbs the filter through schema → service → query.

Marks the parameter as available in both runtimes by dropping the
`[cloud-only]` description prefix and the `x-runtime: [cloud]` tag from
the OpenAPI spec, per the OSS field-drift convention (absent runtime tag
= populated by both local and cloud).

* fix(assets): tighten job_ids — array schema, max_length, narrow except

From cursor-reviews on the parent commit:

- OpenAPI: declare job_ids as `type: array, items: string format: uuid`
  with `style: form, explode: true` so it matches the documented
  contract (and matches sibling include_tags/exclude_tags shape).
  Description now states both accepted shapes explicitly.
- Schema: cap `job_ids` at 500 entries (max_length on the Pydantic
  field) so a client can't splice an unbounded list into the IN clauses.
- Schema: drop `AttributeError` from the except — `raw` only contains
  `str` items by construction, so `uuid.UUID(<str>)` raises `ValueError`
  exclusively; the second clause was dead code.

* fix(assets): tighten job_ids validator + add schema-level tests

Aligns with the parallel hardening from draft PR #13848 (now closed as
a duplicate). The validator now:

- Raises ValueError on non-string list items (was: silently dropped).
- Raises ValueError on non-string / non-list top-level values like dict
  or int (was: silently passed through to Pydantic's downstream coercion).

Adds tests-unit/assets_test/queries/test_list_assets_query.py covering
the validator end-to-end: CSV canonicalization, dedup order, default
empty, invalid UUID, non-string list item, non-string non-list value,
and the max_length=500 boundary.

* feat(prompt): enforce canonical UUID prompt_id at job creation

POST /prompt previously accepted any client-supplied prompt_id verbatim,
str()-coercing even non-strings, and minting the literal job id "None"
for an explicit JSON null. The new GET /api/assets job_ids filter matches
stored job ids as canonical UUIDs exactly, so a non-UUID id minted a job
whose assets could never be filtered.

- validate_job_id (comfy_execution/jobs.py): requires a string in the
  canonical lowercase hyphenated UUID form; raises ValueError otherwise,
  including parseable-but-non-canonical spellings (uppercase, braced, URN,
  bare hex), which would otherwise be silently rewritten and then miss
  every exact-match lookup downstream (history keys, websocket
  correlation, /interrupt, the assets job_ids filter).
- POST /prompt: absent or null prompt_id means the server mints uuid4;
  invalid means 400 invalid_prompt_id on the standard error envelope.
- openapi.yaml: document the request-side prompt_id (format uuid,
  nullable) on PromptRequest.
- tests: unit matrix for validate_job_id; integration tests against the
  booted server covering rejection, acceptance, and null handling.

---------

Co-authored-by: guill <jacob.e.segal@gmail.com>
This commit is contained in:
Matt Miller
2026-06-10 16:55:25 -07:00
committed by GitHub
parent 6d18f4adac
commit e5b7140dcc
11 changed files with 309 additions and 2 deletions

View File

@@ -158,6 +158,56 @@ class TestListReferencesPage:
refs, _, _ = list_references_page(session, sort="name", order="asc")
assert refs[0].name == "large"
def test_job_ids_filter(self, session: Session):
asset = _make_asset(session, "hash1")
job_a = str(uuid.uuid4())
job_b = str(uuid.uuid4())
ref_a = _make_reference(session, asset, name="from_job_a")
ref_a.job_id = job_a
ref_b = _make_reference(session, asset, name="from_job_b")
ref_b.job_id = job_b
_make_reference(session, asset, name="no_job")
session.commit()
# Single job filter
refs, _, total = list_references_page(session, job_ids=[job_a])
assert total == 1
assert refs[0].name == "from_job_a"
# Multi-job filter (IN)
refs, _, total = list_references_page(session, job_ids=[job_a, job_b])
names = sorted(r.name for r in refs)
assert total == 2
assert names == ["from_job_a", "from_job_b"]
# Unknown job id matches nothing
refs, _, total = list_references_page(session, job_ids=[str(uuid.uuid4())])
assert total == 0
assert refs == []
# Empty/None means no filter -> all three references
refs, _, total = list_references_page(session, job_ids=[])
assert total == 3
refs, _, total = list_references_page(session, job_ids=None)
assert total == 3
def test_job_ids_combined_with_other_filters(self, session: Session):
asset = _make_asset(session, "hash1")
job_a = str(uuid.uuid4())
ref_match = _make_reference(session, asset, name="match.bin")
ref_match.job_id = job_a
ref_wrong_name = _make_reference(session, asset, name="other.bin")
ref_wrong_name.job_id = job_a
ref_wrong_job = _make_reference(session, asset, name="match.bin")
ref_wrong_job.job_id = str(uuid.uuid4())
session.commit()
refs, _, total = list_references_page(
session, job_ids=[job_a], name_contains="match"
)
assert total == 1
assert refs[0].id == ref_match.id
class TestFetchReferenceAssetAndTags:
def test_returns_none_for_nonexistent(self, session: Session):

View File

@@ -0,0 +1,60 @@
"""Schema-level unit tests for ListAssetsQuery (no DB required)."""
import uuid
import pytest
from pydantic import ValidationError
from app.assets.api.schemas_in import ListAssetsQuery
class TestJobIdsValidator:
def test_csv_string_parses_and_canonicalizes(self):
a = "AAAAAAAA-BBBB-CCCC-DDDD-EEEEEEEEEEEE"
b = "11111111-2222-3333-4444-555555555555"
q = ListAssetsQuery.model_validate({"job_ids": f"{a},{b}"})
# Canonicalized to lowercase
assert q.job_ids == [a.lower(), b]
def test_repeated_query_params_as_list(self):
a = "11111111-1111-1111-1111-111111111111"
b = "22222222-2222-2222-2222-222222222222"
q = ListAssetsQuery.model_validate({"job_ids": [a, b]})
assert q.job_ids == [a, b]
def test_dedup_preserves_first_seen_order(self):
a = "11111111-1111-1111-1111-111111111111"
b = "22222222-2222-2222-2222-222222222222"
q = ListAssetsQuery.model_validate({"job_ids": [a, b, a]})
assert q.job_ids == [a, b]
def test_default_empty(self):
q = ListAssetsQuery.model_validate({})
assert q.job_ids == []
def test_invalid_uuid_rejected(self):
with pytest.raises(ValidationError) as exc:
ListAssetsQuery.model_validate({"job_ids": "not-a-uuid"})
assert "must be UUIDs" in str(exc.value)
def test_non_string_list_item_rejected(self):
with pytest.raises(ValidationError) as exc:
ListAssetsQuery.model_validate(
{"job_ids": ["11111111-1111-1111-1111-111111111111", 42]}
)
assert "must be strings" in str(exc.value)
def test_non_string_non_list_value_rejected(self):
with pytest.raises(ValidationError) as exc:
ListAssetsQuery.model_validate({"job_ids": {"bad": "shape"}})
assert "must be a string or list of strings" in str(exc.value)
def test_max_length_enforced(self):
too_many = [str(uuid.uuid4()) for _ in range(501)]
with pytest.raises(ValidationError) as exc:
ListAssetsQuery.model_validate({"job_ids": too_many})
assert exc.value.errors()[0]["type"] == "too_long"
def test_max_length_boundary_accepted(self):
at_cap = [str(uuid.uuid4()) for _ in range(500)]
q = ListAssetsQuery.model_validate({"job_ids": at_cap})
assert len(q.job_ids) == 500

View File

@@ -0,0 +1,69 @@
"""POST /prompt enforces canonical-UUID job ids at creation time.
Lives in assets_test because it uses this suite's booted-server fixture and
because the invariant exists for the assets pipeline: the GET /api/assets
``job_ids`` filter matches stored job ids exactly, so a job minted with a
non-canonical id would produce assets the filter can never find.
The prompt bodies here are intentionally invalid workflows — prompt_id
validation happens before workflow validation, so a rejected id returns
``invalid_prompt_id`` while an accepted id falls through to the ordinary
workflow-validation error (proving it cleared the id check).
"""
import requests
def _post_prompt(http: requests.Session, api_base: str, body: dict) -> requests.Response:
return http.post(api_base + "/prompt", json=body, timeout=30)
def _error_type(r: requests.Response) -> str:
return r.json()["error"]["type"]
def test_non_uuid_prompt_id_rejected(http: requests.Session, api_base: str):
r = _post_prompt(http, api_base, {"prompt": {}, "prompt_id": "not-a-uuid"})
assert r.status_code == 400, r.text
assert _error_type(r) == "invalid_prompt_id"
def test_non_string_prompt_id_rejected(http: requests.Session, api_base: str):
# Previously str()-coerced (123 became the job id "123"); must now be a 400,
# not a 500 from uuid.UUID choking on a non-string.
r = _post_prompt(http, api_base, {"prompt": {}, "prompt_id": 123})
assert r.status_code == 400, r.text
assert _error_type(r) == "invalid_prompt_id"
def test_non_canonical_uuid_rejected(http: requests.Session, api_base: str):
# Parseable as a UUID, but not the canonical lowercase form: rejected
# loudly rather than silently rewritten (downstream lookups match the
# stored id exactly).
r = _post_prompt(
http,
api_base,
{"prompt": {}, "prompt_id": "AAAAAAAA-BBBB-4CCC-8DDD-EEEEEEEEEEEE"},
)
assert r.status_code == 400, r.text
assert _error_type(r) == "invalid_prompt_id"
def test_canonical_uuid_accepted(http: requests.Session, api_base: str):
# The id clears validation; the empty workflow then fails ordinary prompt
# validation, proving the request got past the id check.
r = _post_prompt(
http,
api_base,
{"prompt": {}, "prompt_id": "aaaaaaaa-bbbb-4ccc-8ddd-eeeeeeeeeeee"},
)
assert r.status_code == 400, r.text
assert _error_type(r) != "invalid_prompt_id"
def test_null_prompt_id_not_rejected(http: requests.Session, api_base: str):
# Explicit null means "server generates" and must not be rejected as an
# invalid id. (The minted id itself is not observable here because the
# workflow is invalid; unit tests cover validate_job_id directly.)
r = _post_prompt(http, api_base, {"prompt": {}, "prompt_id": None})
assert r.status_code == 400, r.text
assert _error_type(r) != "invalid_prompt_id"