peter/siftlode
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

test(migrations)+fix: widen drift check to column shape; notifications.created_at NOT NULL (S2 review #2)

Browse Source 
The drift check dropped every diff except table/column presence, so a column
type or nullability change would slip through. Add modify_type/modify_nullable
to the ops it fails on. That immediately surfaced a real pre-existing drift:
notifications.created_at was created NULLABLE by migration 0017 but the model
has always declared it NOT NULL (Mapped[datetime] + server_default=func.now()).
Migration 0066 enforces NOT NULL to match the model — safe, since the
server_default has filled every row since day one.
This commit is contained in:
peter
2026-08-18 01:39:46 +02:00
parent e764ca4568
commit 472cefed55
2 changed files with 35 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files
backend/tests/test_migrations.py
+11 -7
View File
@@ -9,11 +9,12 @@ Two things the prod boot never checks:
Both use the `db` fixture, which has already brought the schema to head via `_schema`.
Scope note: we assert on STRUCTURAL drift only — tables and columns present in the model but not the
schema, or vice versa. `compare_metadata` also reports index/constraint/sequence/type differences, but
those are dominated by migration-only objects (partial/expression/GIN indexes, CHECK constraints, SERIAL
sequences) that can't be expressed on `Base.metadata`, so comparing them wholesale is noise. Missing/extra
columns and tables — the "someone changed a model without a migration" case — are the signal worth gating.
Scope note: we gate on COLUMN/TABLE drift — a table or column present on one side and not the other, or
a column whose type/nullability disagrees. `compare_metadata` also reports index/constraint/sequence
differences, but those are dominated by migration-only objects (partial/expression/GIN indexes, CHECK
constraints, SERIAL sequences) that can't be expressed on `Base.metadata`, so comparing them wholesale is
noise; they're filtered out. The column/table signal is the "someone changed a model without a migration"
case worth failing on.
"""
from alembic import command
from alembic.autogenerate import compare_metadata
@@ -22,8 +23,11 @@ from alembic.migration import MigrationContext
from app.db import Base, engine
# The diff ops that mean a real table/column is present on one side and not the other.
_DRIFT_OPS = {"add_table", "remove_table", "add_column", "remove_column"}
# The diff ops that mean a real table/column disagreement (presence AND shape) — not the index/
# constraint/sequence noise from migration-only objects.
_DRIFT_OPS = {
"add_table", "remove_table", "add_column", "remove_column", "modify_type", "modify_nullable",
}
def _structural_drift():