brain-of-reese/tests/integration/test_migration_0007.py

"""Integration: migration 0007 (git_sources.kind + path) schema contract.

Drives the **real Alembic engine** against the live dev database
(``podman compose up -d db``), mirroring the style of
``test_migration_0004.py`` / ``test_migration_0006.py``
(information_schema assertions on the state the migration must leave).
The tests target revision ``0007`` explicitly so later migrations
cannot break them:

* upgrade 0006 → 0007 → ``git_sources`` gains ``kind TEXT NOT NULL``
  (server default ``'git'``, check constraint ``ck_git_sources_kind``:
  ``kind IN ('git', 'local')``) and ``path TEXT`` (nullable) with the
  unique index ``uq_git_sources_path``; a row inserted before the
  upgrade (the pre-0007 insert shape) keeps ``kind='git'`` /
  ``path=NULL`` after it;
* the check constraint rejects any kind other than ``git``/``local``;
* the unique index rejects duplicate local paths but tolerates NULL
  paths (git rows);
* downgrade to 0006 → both columns, the constraint, and the index are
  gone;
* upgrade back to 0007 → they are back (round-trip).

The ``alembic`` fixture guarantees the DB ends at head even if a test
fails or the process is interrupted.
"""
from __future__ import annotations

from collections.abc import Iterator
from typing import Any

import pytest
from alembic.config import Config
from sqlalchemy import text
from sqlalchemy.exc import IntegrityError
from sqlalchemy.orm import Session

from alembic import command
from app.db import db_available

URL_BASE = "https://git.example.com/mig0007"
PATH_BASE = "/tmp/brain-of-reese-mig0007"


@pytest.fixture()
def alembic(db: Session) -> Iterator[Config]:
    """Real Alembic config bound to the dev DB (URL from app settings).

    Starts at head (repairs an interrupted earlier run); teardown upgrades
    to head no matter what happened, so the dev DB is never left below
    head.
    """
    if not db_available():
        pytest.skip("Postgres not reachable — run `podman compose up -d db` first")
    cfg = Config()  # no alembic.ini file — env.py gets the URL from app config
    cfg.set_main_option("script_location", "alembic")
    command.upgrade(cfg, "head")
    try:
        yield cfg
    finally:
        # Release the test session's open transaction BEFORE the repair
        # DDL: psycopg3 keeps its snapshot alive while idle-in-transaction,
        # and 0029's CREATE INDEX CONCURRENTLY phase 3 waits for every
        # older-snapshot transaction — an unreleased session would park
        # the build forever.
        db.rollback()
        command.upgrade(cfg, "head")


def _column(db: Session, column: str) -> tuple[Any, ...] | None:
    """(data_type, is_nullable, column_default) for one git_sources column."""
    row = db.execute(
        text(
            "SELECT data_type, is_nullable, column_default"
            " FROM information_schema.columns"
            " WHERE table_name = 'git_sources' AND column_name = :c"
        ),
        {"c": column},
    ).fetchone()
    return tuple(row) if row is not None else None


def _version(db: Session) -> str | None:
    return db.execute(text("SELECT version_num FROM alembic_version")).scalar()


def _check_constraint_def(db: Session) -> str | None:
    """Definition of ``ck_git_sources_kind``, or None if it does not exist.

    Only call while ``git_sources`` exists (the ``::regclass`` cast errors
    otherwise).
    """
    row = db.execute(
        text(
            "SELECT pg_get_constraintdef(oid) FROM pg_constraint"
            " WHERE conname = 'ck_git_sources_kind'"
            " AND conrelid = 'git_sources'::regclass"
        )
    ).fetchone()
    return row[0] if row is not None else None


def _unique_path_index(db: Session) -> int:
    """1 iff ``uq_git_sources_path`` exists as a UNIQUE index."""
    count: Any = db.execute(
        text(
            "SELECT count(*) FROM pg_indexes"
            " WHERE tablename = 'git_sources' AND indexname = 'uq_git_sources_path'"
            " AND indexdef ILIKE 'CREATE UNIQUE%'"
        )
    ).scalar()
    assert count is not None, "pg_indexes count must be an int"
    return int(count)


def _insert(db: Session, url: str, kind: str | None = None, path: str | None = None) -> None:
    """Insert one git_sources row; kind/path omitted → pre-0007 shape."""
    if kind is None and path is None:
        db.execute(
            text("INSERT INTO git_sources (id, url) VALUES (gen_random_uuid(), :u)"),
            {"u": url},
        )
    else:
        db.execute(
            text(
                "INSERT INTO git_sources (id, url, kind, path)"
                " VALUES (gen_random_uuid(), :u, :k, :p)"
            ),
            {"u": url, "k": kind, "p": path},
        )
    db.commit()


def _delete_by_url(db: Session, url: str) -> None:
    db.execute(text("DELETE FROM git_sources WHERE url = :u"), {"u": url})
    db.commit()


def test_upgrade_to_0007_adds_kind_and_path(db: Session, alembic: Config) -> None:
    """Upgrade 0006 → 0007: both columns exist with the locked types,
    nullability, and defaults, plus the CHECK constraint and the unique
    path index."""
    command.downgrade(alembic, "0006")  # start from the pre-0007 state
    assert _version(db) == "0006"

    command.upgrade(alembic, "0007")
    assert _version(db) == "0007", "alembic_version must be at 0007"

    kind = _column(db, "kind")
    assert kind is not None, "git_sources.kind is missing"
    assert kind[0] == "text", "git_sources.kind must be TEXT"
    assert kind[1] == "NO", "git_sources.kind must be NOT NULL"
    assert kind[2] is not None and "'git'" in kind[2], (
        "git_sources.kind must have server default 'git'"
    )

    path = _column(db, "path")
    assert path is not None, "git_sources.path is missing"
    assert path[0] == "text", "git_sources.path must be TEXT"
    assert path[1] == "YES", "git_sources.path must be NULLABLE"

    constraint = _check_constraint_def(db)
    assert constraint is not None, "ck_git_sources_kind is missing"
    assert "git" in constraint and "local" in constraint, (
        f"ck_git_sources_kind must restrict kind to git|local, got: {constraint}"
    )

    assert _unique_path_index(db) == 1, "uq_git_sources_path unique index is missing"


def test_pre_0007_row_reads_as_git(db: Session, alembic: Config) -> None:
    """A row inserted before the upgrade (url only — the pre-0007 insert
    shape) reads as ``kind='git'``, ``path=NULL`` after it."""
    command.downgrade(alembic, "0006")
    url = f"{URL_BASE}/pre-existing.git"
    _insert(db, url)  # no kind/path columns exist at 0006
    try:
        command.upgrade(alembic, "0007")
        kind, path = db.execute(
            text("SELECT kind, path FROM git_sources WHERE url = :u"), {"u": url}
        ).one()
        assert kind == "git", "a pre-0007 row must read as kind='git'"
        assert path is None, "a pre-0007 row must keep path=NULL"
    finally:
        _delete_by_url(db, url)


def test_kind_defaults_to_git_for_new_inserts(db: Session, alembic: Config) -> None:
    """An insert that omits kind (the API's pre-phase-38 shape) lands as
    ``kind='git'`` via the server default."""
    command.upgrade(alembic, "head")
    url = f"{URL_BASE}/default-kind.git"
    _insert(db, url)
    try:
        kind, path = db.execute(
            text("SELECT kind, path FROM git_sources WHERE url = :u"), {"u": url}
        ).one()
        assert kind == "git", "git_sources.kind must default to 'git'"
        assert path is None, "git_sources.path must default to NULL"
    finally:
        _delete_by_url(db, url)


def test_check_constraint_rejects_unknown_kind(db: Session, alembic: Config) -> None:
    """``ck_git_sources_kind`` is what later API validation relies on:
    any kind other than git|local raises IntegrityError."""
    command.upgrade(alembic, "head")
    with pytest.raises(IntegrityError):
        _insert(db, f"{URL_BASE}/bogus-kind.git", kind="bogus")
    db.rollback()  # the IntegrityError aborts the open transaction


def test_duplicate_local_path_rejected(db: Session, alembic: Config) -> None:
    """The unique path index is what the API's 409 relies on: two local
    rows with the same path are rejected (NULL paths stay distinct —
    git rows are unaffected)."""
    command.upgrade(alembic, "head")
    url_a = f"{URL_BASE}/dup-path-a.git"
    url_b = f"{URL_BASE}/dup-path-b.git"
    url_c = f"{URL_BASE}/dup-path-c.git"
    path = f"{PATH_BASE}/shared"
    try:
        _insert(db, url_a, kind="local", path=path)
        with pytest.raises(IntegrityError):
            _insert(db, url_b, kind="local", path=path)
        db.rollback()
        # NULL paths are distinct under the unique index (git rows).
        _insert(db, url_b)
        _insert(db, url_c)
    finally:
        db.rollback()
        _delete_by_url(db, url_a)
        _delete_by_url(db, url_b)
        _delete_by_url(db, url_c)


def test_downgrade_to_0006_drops_columns(db: Session, alembic: Config) -> None:
    """Downgrade to 0006: both columns, the CHECK constraint, and the
    unique index are dropped (A13 — reversible)."""
    command.downgrade(alembic, "0006")
    assert _version(db) == "0006"

    assert _column(db, "kind") is None, "git_sources.kind must be dropped"
    assert _column(db, "path") is None, "git_sources.path must be dropped"
    assert _check_constraint_def(db) is None, "ck_git_sources_kind must be dropped"
    assert _unique_path_index(db) == 0, "uq_git_sources_path must be dropped"


def test_upgrade_round_trip_restores_columns(db: Session, alembic: Config) -> None:
    """Downgrade to 0006, then upgrade back to 0007: columns, default,
    constraint, and index are back."""
    command.downgrade(alembic, "0006")
    command.upgrade(alembic, "0007")
    assert _version(db) == "0007", "round-trip upgrade must land at 0007"

    kind = _column(db, "kind")
    assert kind is not None, "git_sources.kind must be back"
    assert kind[1] == "NO" and kind[2] is not None and "'git'" in kind[2], (
        "git_sources.kind must keep its NOT NULL 'git' default after the round-trip"
    )

    path = _column(db, "path")
    assert path is not None and path[1] == "YES", "git_sources.path must be back"

    constraint = _check_constraint_def(db)
    assert constraint is not None, "ck_git_sources_kind must be back"

    assert _unique_path_index(db) == 1, "uq_git_sources_path must be back"