YoVinchen
ecdb10c37a
Some checks failed
verify / verify (push) Has been cancelled
Separate local parsing from model indexing, bind review decisions to immutable manifests, persist vectors behind active profiles, and expose retrieval, chat, evaluation, and document workflows through the React workbench. Constraint: Live Bailian authentication currently fails for all three configured capabilities Rejected: Direct upload-to-embedding flow | bypasses local review and manifest binding Confidence: high Scope-risk: broad Directive: Keep private-data deployment blocked until authentication, RBAC, and separate database roles land Tested: make verify; fresh and replay Docker document smoke; worker recovery smoke; frozen synthetic evaluation; migration 0003-0004 roundtrip Not-tested: Successful live Bailian calls, OCR, real multi-user authorization
112 lines
4.5 KiB
Python
112 lines
4.5 KiB
Python
from __future__ import annotations
|
|
|
|
import re
|
|
from pathlib import Path
|
|
|
|
ROOT = Path(__file__).resolve().parents[3]
|
|
MIGRATION_PATH = ROOT / "backend/migrations/versions/0003_document_ingestion.py"
|
|
MIGRATION = MIGRATION_PATH.read_text(encoding="utf-8")
|
|
NORMALIZED = " ".join(MIGRATION.lower().split())
|
|
|
|
|
|
def _table(name: str) -> str:
|
|
pattern = re.compile(rf"(?ms)create table rag\.{re.escape(name)} \((.*?)^ \);?")
|
|
match = pattern.search(MIGRATION.lower())
|
|
assert match is not None, f"missing table rag.{name}"
|
|
return " ".join(match.group(1).split())
|
|
|
|
|
|
def test_revision_is_additive_after_model_profiles() -> None:
|
|
assert 'revision: str = "0003_document_ingestion"' in MIGRATION
|
|
assert 'down_revision: str | none = "0002_model_profiles"' in MIGRATION.lower()
|
|
assert "alter table rag.documents" not in NORMALIZED
|
|
assert "alter table rag.chunks" not in NORMALIZED
|
|
assert "drop table if exists rag.documents" not in NORMALIZED
|
|
assert "drop table if exists rag.background_jobs" not in NORMALIZED
|
|
|
|
|
|
def test_uploads_bind_actor_scope_idempotency_content_and_completion() -> None:
|
|
table = _table("document_uploads")
|
|
|
|
for column in (
|
|
"actor_subject text not null",
|
|
"knowledge_base_id uuid not null",
|
|
"access_scope_id uuid not null",
|
|
"idempotency_key_hash char(64) not null",
|
|
"request_fingerprint char(64) not null",
|
|
"expected_size bigint not null",
|
|
"expected_sha256 char(64) not null",
|
|
"storage_key uuid not null",
|
|
"status text not null default 'created'",
|
|
"document_id uuid",
|
|
"parse_job_id uuid",
|
|
):
|
|
assert column in table
|
|
assert "foreign key (knowledge_base_id, access_scope_id)" in table
|
|
assert "references rag.access_scopes (knowledge_base_id, id)" in table
|
|
assert "unique (actor_subject, idempotency_key_hash)" in table
|
|
assert "unique (storage_key)" in table
|
|
assert "status in ('created', 'stored', 'completed')" in table
|
|
assert "actual_size = expected_size" in table
|
|
assert "actual_sha256 = expected_sha256" in table
|
|
assert "status = 'completed'" in table
|
|
assert "document_id is not null" in table
|
|
assert "parse_job_id is not null" in table
|
|
assert "completed_at is not null" in table
|
|
assert "original_filename !~ '[/\\\\\\\\]'" in table
|
|
|
|
|
|
def test_upload_audit_is_append_only_metadata_without_sensitive_fields() -> None:
|
|
table = _table("document_upload_events")
|
|
|
|
assert "upload_id uuid not null" in table
|
|
assert "actor_subject text not null" in table
|
|
assert "trace_id uuid not null" in table
|
|
assert "event_type in ('created', 'stored', 'completed')" in table
|
|
assert "jsonb_typeof(metadata) = 'object'" in table
|
|
assert "metadata_has_no_credentials" in table
|
|
for forbidden in (
|
|
"api[_-]?key",
|
|
"secret",
|
|
"password",
|
|
"token",
|
|
"authorization",
|
|
"credential",
|
|
"storage[_-]?key",
|
|
"path",
|
|
):
|
|
assert forbidden in table
|
|
|
|
|
|
def test_pages_and_blocks_preserve_version_source_anchors() -> None:
|
|
pages = _table("document_pages")
|
|
blocks = _table("document_blocks")
|
|
|
|
assert "foreign key (document_version_id)" in pages
|
|
assert "references rag.document_versions (id) on delete cascade" in pages
|
|
assert "unique (document_version_id, ordinal)" in pages
|
|
assert "page_number is null or page_number > 0" in pages
|
|
assert "text_sha256 ~ '^[0-9a-f]{64}$'" in pages
|
|
assert "line_start > 0 and line_end >= line_start" in pages
|
|
|
|
assert "foreign key (document_version_id, page_id)" in blocks
|
|
assert "references rag.document_pages (document_version_id, id)" in blocks
|
|
assert "block_kind in ('heading', 'paragraph', 'table_row')" in blocks
|
|
assert "jsonb_typeof(section_path) = 'array'" in blocks
|
|
assert "anchor_id ~ '^[0-9a-f]{64}$'" in blocks
|
|
assert "normalized_text_sha256 ~ '^[0-9a-f]{64}$'" in blocks
|
|
assert "char_end > char_start" in blocks
|
|
assert "line_end >= line_start" in blocks
|
|
assert "unique (document_version_id, anchor_id)" in blocks
|
|
|
|
|
|
def test_downgrade_drops_only_new_dependents_in_safe_order() -> None:
|
|
block = NORMALIZED.index("drop table if exists rag.document_blocks")
|
|
page = NORMALIZED.index("drop table if exists rag.document_pages")
|
|
event = NORMALIZED.index("drop table if exists rag.document_upload_events")
|
|
upload = NORMALIZED.index("drop table if exists rag.document_uploads")
|
|
|
|
assert block < page < event < upload
|
|
assert "drop table if exists rag.documents" not in NORMALIZED
|
|
assert "drop table if exists rag.document_versions" not in NORMALIZED
|