fe346d983b
Repeated knowledge-ingest of identical content (same sha256, new asset_id -> new canonical object key) reuses the Document row but appends a second ORIGINAL_PDF artifact, because ensure_artifact identity is (document_id, storage_key). process_document_id then crashed with MultipleResultsFound on scalar_one_or_none and Celery retried forever. New latest_artifact helper picks the newest row deterministically (created_at DESC, id DESC) — every duplicate references byte-identical objects (sha256 verified at ingest), so the latest reference is always safe. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
94 lines
3.1 KiB
Python
94 lines
3.1 KiB
Python
"""G6: ORIGINAL_PDF artifact lookup must tolerate duplicate rows.
|
|
|
|
Repeated knowledge-ingest of identical content (same sha256, new asset_id →
|
|
new canonical object key) reuses the Document row but appends a second
|
|
ORIGINAL_PDF artifact (ensure_artifact identity is document_id+storage_key).
|
|
The old ``scalar_one_or_none()`` in ``process_document_id`` then raised
|
|
``MultipleResultsFound`` → Celery retry storm. The lookup must instead pick
|
|
the newest artifact deterministically — any row is byte-identical (sha256
|
|
verified at ingest), so the latest reference is always safe.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import uuid
|
|
|
|
from app.db.models import ArtifactType, DocumentArtifact
|
|
from app.storage.artifacts import latest_artifact
|
|
|
|
|
|
class _FakeScalars:
|
|
def __init__(self, rows):
|
|
self._rows = rows
|
|
|
|
def first(self):
|
|
return self._rows[0] if self._rows else None
|
|
|
|
|
|
class _FakeResult:
|
|
def __init__(self, rows):
|
|
self._rows = rows
|
|
|
|
def scalars(self):
|
|
return _FakeScalars(self._rows)
|
|
|
|
|
|
class _FakeSession:
|
|
def __init__(self, rows):
|
|
self.rows = rows
|
|
self.statements = []
|
|
|
|
def execute(self, stmt):
|
|
self.statements.append(stmt)
|
|
return _FakeResult(self.rows)
|
|
|
|
|
|
def _artifact(key: str) -> DocumentArtifact:
|
|
return DocumentArtifact(
|
|
id=uuid.uuid4(),
|
|
document_id=uuid.uuid4(),
|
|
artifact_type=ArtifactType.ORIGINAL_PDF,
|
|
storage_bucket="teamhub-mailhub-originals",
|
|
storage_key=key,
|
|
)
|
|
|
|
|
|
def test_latest_artifact_returns_first_row_without_raising():
|
|
rows = [_artifact("mailhub/2026/07/11/b/original/x.pdf"),
|
|
_artifact("mailhub/2026/07/11/a/original/x.pdf")]
|
|
db = _FakeSession(rows)
|
|
got = latest_artifact(db, document_id=rows[0].document_id,
|
|
artifact_type=ArtifactType.ORIGINAL_PDF)
|
|
assert got is rows[0]
|
|
|
|
|
|
def test_latest_artifact_none_when_absent():
|
|
db = _FakeSession([])
|
|
assert latest_artifact(db, document_id=uuid.uuid4(),
|
|
artifact_type=ArtifactType.ORIGINAL_PDF) is None
|
|
|
|
|
|
def test_latest_artifact_query_is_deterministic_newest_first():
|
|
"""The emitted SQL must filter by document_id+artifact_type and order
|
|
newest-first with a stable tiebreaker — that is what makes the pick
|
|
deterministic when duplicates exist."""
|
|
db = _FakeSession([])
|
|
latest_artifact(db, document_id=uuid.uuid4(), artifact_type=ArtifactType.ORIGINAL_PDF)
|
|
(stmt,) = db.statements
|
|
sql = str(stmt.compile(compile_kwargs={"literal_binds": False}))
|
|
assert "document_artifacts.document_id" in sql
|
|
assert "document_artifacts.artifact_type" in sql
|
|
assert "ORDER BY document_artifacts.created_at DESC, document_artifacts.id DESC" in sql
|
|
|
|
|
|
def test_pipeline_uses_tolerant_lookup():
|
|
"""process_document_id must not use scalar_one_or_none for the ORIGINAL_PDF
|
|
lookup (the G6 crash); it goes through latest_artifact instead."""
|
|
import inspect
|
|
|
|
from app.ingestion import pipeline
|
|
|
|
src = inspect.getsource(pipeline.process_document_id)
|
|
assert "latest_artifact" in src
|
|
assert "scalar_one_or_none" not in src
|