Source code for cardinal_pythonlib.sqlalchemy.alembic_func

#!/usr/bin/env python
# cardinal_pythonlib/sqlalchemy/alembic_func.py

"""
===============================================================================

    Original code copyright (C) 2009-2022 Rudolf Cardinal (rudolf@pobox.com).

    This file is part of cardinal_pythonlib.

    Licensed under the Apache License, Version 2.0 (the "License");
    you may not use this file except in compliance with the License.
    You may obtain a copy of the License at

        https://www.apache.org/licenses/LICENSE-2.0

    Unless required by applicable law or agreed to in writing, software
    distributed under the License is distributed on an "AS IS" BASIS,
    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    See the License for the specific language governing permissions and
    limitations under the License.

===============================================================================

**Support functions for Alembic, the migration tool for SQLAlchemy.**

"""

import os
import re
import subprocess
from typing import Tuple

from alembic.config import Config
from alembic.util.exc import CommandError
from alembic.runtime.migration import MigrationContext
from alembic.runtime.environment import EnvironmentContext
from alembic.script import ScriptDirectory
from sqlalchemy.engine import create_engine

from cardinal_pythonlib.fileops import preserve_cwd
from cardinal_pythonlib.logs import get_brace_style_log_with_null_handler

log = get_brace_style_log_with_null_handler(__name__)


# =============================================================================
# Constants for Alembic
# =============================================================================
# https://alembic.readthedocs.org/en/latest/naming.html
# http://docs.sqlalchemy.org/en/latest/core/constraints.html#configuring-constraint-naming-conventions  # noqa

ALEMBIC_NAMING_CONVENTION = {
    "ix": "ix_%(column_0_label)s",
    "uq": "uq_%(table_name)s_%(column_0_name)s",
    # "ck": "ck_%(table_name)s_%(constraint_name)s",  # too long?
    # ... https://groups.google.com/forum/#!topic/sqlalchemy/SIT4D8S9dUg
    "ck": "ck_%(table_name)s_%(column_0_name)s",
    "fk": "fk_%(table_name)s_%(column_0_name)s_%(referred_table_name)s",
    "pk": "pk_%(table_name)s",
}

DEFAULT_ALEMBIC_VERSION_TABLE = "alembic_version"


# =============================================================================
# Alembic revision/migration system
# =============================================================================
# https://stackoverflow.com/questions/24622170/using-alembic-api-from-inside-application-code  # noqa


[docs]def get_head_revision_from_alembic( alembic_config_filename: str, alembic_base_dir: str = None, version_table: str = DEFAULT_ALEMBIC_VERSION_TABLE, ) -> str: """ Ask Alembic what its head revision is (i.e. where the Python code would like the database to be at). Arguments: alembic_config_filename: config filename alembic_base_dir: directory to start in, so relative paths in the config file work. version_table: table name for Alembic versions """ if alembic_base_dir is None: alembic_base_dir = os.path.dirname(alembic_config_filename) os.chdir(alembic_base_dir) # so the directory in the config file works config = Config(alembic_config_filename) script = ScriptDirectory.from_config(config) with EnvironmentContext(config, script, version_table=version_table): return script.get_current_head()
[docs]def get_current_revision( database_url: str, version_table: str = DEFAULT_ALEMBIC_VERSION_TABLE ) -> str: """ Ask the database what its current revision is. Arguments: database_url: SQLAlchemy URL for the database version_table: table name for Alembic versions """ engine = create_engine(database_url) conn = engine.connect() opts = {"version_table": version_table} mig_context = MigrationContext.configure(conn, opts=opts) return mig_context.get_current_revision()
[docs]def get_current_and_head_revision( database_url: str, alembic_config_filename: str, alembic_base_dir: str = None, version_table: str = DEFAULT_ALEMBIC_VERSION_TABLE, ) -> Tuple[str, str]: """ Returns a tuple of ``(current_revision, head_revision)``; see :func:`get_current_revision` and :func:`get_head_revision_from_alembic`. Arguments: database_url: SQLAlchemy URL for the database alembic_config_filename: config filename alembic_base_dir: directory to start in, so relative paths in the config file work. version_table: table name for Alembic versions """ # Where we are head_revision = get_head_revision_from_alembic( alembic_config_filename=alembic_config_filename, alembic_base_dir=alembic_base_dir, version_table=version_table, ) log.debug("Intended database version: {}", head_revision) # Where we want to be current_revision = get_current_revision( database_url=database_url, version_table=version_table ) log.debug("Current database version: {}", current_revision) # Are we where we want to be? return current_revision, head_revision
@preserve_cwd def upgrade_database( alembic_config_filename: str, alembic_base_dir: str = None, starting_revision: str = None, destination_revision: str = "head", version_table: str = DEFAULT_ALEMBIC_VERSION_TABLE, as_sql: bool = False, ) -> None: """ Use Alembic to upgrade our database. See https://alembic.readthedocs.org/en/latest/api/runtime.html but also, in particular, ``site-packages/alembic/command.py`` Arguments: alembic_config_filename: config filename alembic_base_dir: directory to start in, so relative paths in the config file work starting_revision: revision to start at (typically ``None`` to ask the database) destination_revision: revision to aim for (typically ``"head"`` to migrate to the latest structure) version_table: table name for Alembic versions as_sql: run in "offline" mode: print the migration SQL, rather than modifying the database. See https://alembic.zzzcomputing.com/en/latest/offline.html """ if alembic_base_dir is None: alembic_base_dir = os.path.dirname(alembic_config_filename) os.chdir(alembic_base_dir) # so the directory in the config file works config = Config(alembic_config_filename) script = ScriptDirectory.from_config(config) # noinspection PyUnusedLocal,PyProtectedMember def upgrade(rev, context): return script._upgrade_revs(destination_revision, rev) log.info( "Upgrading database to revision {!r} using Alembic", destination_revision, ) with EnvironmentContext( config, script, fn=upgrade, as_sql=as_sql, starting_rev=starting_revision, destination_rev=destination_revision, tag=None, version_table=version_table, ): script.run_env() log.info("Database upgrade completed") @preserve_cwd def downgrade_database( alembic_config_filename: str, destination_revision: str, alembic_base_dir: str = None, starting_revision: str = None, version_table: str = DEFAULT_ALEMBIC_VERSION_TABLE, as_sql: bool = False, ) -> None: """ Use Alembic to downgrade our database. USE WITH EXTREME CAUTION. "revision" is the destination revision. See https://alembic.readthedocs.org/en/latest/api/runtime.html but also, in particular, ``site-packages/alembic/command.py`` Arguments: alembic_config_filename: config filename alembic_base_dir: directory to start in, so relative paths in the config file work starting_revision: revision to start at (typically ``None`` to ask the database) destination_revision: revision to aim for version_table: table name for Alembic versions as_sql: run in "offline" mode: print the migration SQL, rather than modifying the database. See https://alembic.zzzcomputing.com/en/latest/offline.html """ if alembic_base_dir is None: alembic_base_dir = os.path.dirname(alembic_config_filename) os.chdir(alembic_base_dir) # so the directory in the config file works config = Config(alembic_config_filename) script = ScriptDirectory.from_config(config) # noinspection PyUnusedLocal,PyProtectedMember def downgrade(rev, context): return script._downgrade_revs(destination_revision, rev) log.info( "Downgrading database to revision {!r} using Alembic", destination_revision, ) with EnvironmentContext( config, script, fn=downgrade, as_sql=as_sql, starting_rev=starting_revision, destination_rev=destination_revision, tag=None, version_table=version_table, ): script.run_env() log.info("Database downgrade completed") @preserve_cwd def create_database_migration_numbered_style( alembic_ini_file: str, alembic_versions_dir: str, message: str, n_sequence_chars: int = 4, ) -> None: """ Create a new Alembic migration script. Alembic compares the **state of the database** to the **state of the metadata**, and generates a migration that brings the former up to the latter. (It does **not** compare the most recent revision to the current metadata, so make sure your database is up to date with the most recent revision before running this!) You **must check** that the autogenerated revisions are sensible. How does it know where to look for the database? 1. This function changes into the directory of the Alembic ``.ini`` file and calls the external program .. code-block:: bash alembic -c ALEMBIC_INI_FILE revision --autogenerate -m MESSAGE --rev-id REVISION_ID 2. The Alembic ``.ini`` file points (via the ``script_location`` variable) to a directory containing your ``env.py``. Alembic loads this script. 3. That script typically works out the database URL and calls further into the Alembic code. See https://alembic.zzzcomputing.com/en/latest/autogenerate.html. Regarding filenames: the default ``n_sequence_chars`` of 4 is like Django and gives files with names like .. code-block:: none 0001_x.py, 0002_y.py, ... NOTE THAT TO USE A NON-STANDARD ALEMBIC VERSION TABLE, YOU MUST SPECIFY THAT IN YOUR ``env.py`` (see e.g. CamCOPS). Args: alembic_ini_file: filename of Alembic ``alembic.ini`` file alembic_versions_dir: directory in which you keep your Python scripts, one per Alembic revision message: message to be associated with this revision n_sequence_chars: number of numerical sequence characters to use in the filename/revision (see above). """ # noqa file_regex = r"\d{" + str(n_sequence_chars) + r"}_\S*\.py$" _, _, existing_version_filenames = next( os.walk(alembic_versions_dir), (None, None, []) ) existing_version_filenames = [ x for x in existing_version_filenames if re.match(file_regex, x) ] log.debug( "Existing Alembic version script filenames: {!r}", existing_version_filenames, ) current_seq_strs = [ x[:n_sequence_chars] for x in existing_version_filenames ] current_seq_strs.sort() if not current_seq_strs: current_seq_str = None new_seq_no = 1 else: current_seq_str = current_seq_strs[-1] new_seq_no = max(int(x) for x in current_seq_strs) + 1 new_seq_str = str(new_seq_no).zfill(n_sequence_chars) log.info( """ Generating new revision with Alembic... Last revision was: {} New revision will be: {} [If it fails with "Can't locate revision identified by...", you might need to DROP the Alembic version table (by default named 'alembic_version', but you may have elected to change that in your env.py.] """, current_seq_str, new_seq_str, ) alembic_ini_dir = os.path.dirname(alembic_ini_file) os.chdir(alembic_ini_dir) cmdargs = [ "alembic", "-c", alembic_ini_file, "revision", "--autogenerate", "-m", message, "--rev-id", new_seq_str, ] log.info("From directory {!r}, calling: {!r}", alembic_ini_dir, cmdargs) subprocess.call(cmdargs)
[docs]def stamp_allowing_unusual_version_table( config: Config, revision: str, sql: bool = False, tag: str = None, version_table: str = DEFAULT_ALEMBIC_VERSION_TABLE, ) -> None: """ Stamps the Alembic version table with the given revision; don't run any migrations. This function is a clone of ``alembic.command.stamp()``, but allowing ``version_table`` to change. See https://alembic.zzzcomputing.com/en/latest/api/commands.html#alembic.command.stamp """ # noqa script = ScriptDirectory.from_config(config) starting_rev = None if ":" in revision: if not sql: raise CommandError("Range revision not allowed") starting_rev, revision = revision.split(":", 2) # noinspection PyUnusedLocal def do_stamp(rev: str, context): # noinspection PyProtectedMember return script._stamp_revs(revision, rev) with EnvironmentContext( config, script, fn=do_stamp, as_sql=sql, destination_rev=revision, starting_rev=starting_rev, tag=tag, version_table=version_table, ): script.run_env()