##############################################################################
#
# Copyright (c) 2001, 2002 Zope Foundation and Contributors.
# All Rights Reserved.
#
# This software is subject to the provisions of the Zope Public License,
# Version 2.1 (ZPL). A copy of the ZPL should accompany this distribution.
# THIS SOFTWARE IS PROVIDED "AS IS" AND ANY AND ALL EXPRESS OR IMPLIED
# WARRANTIES ARE DISCLAIMED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
# WARRANTIES OF TITLE, MERCHANTABILITY, AGAINST INFRINGEMENT, AND FITNESS
# FOR A PARTICULAR PURPOSE.
#
##############################################################################
from zope.interface import Attribute
from zope.interface import Interface
[docs]class ITransactionManager(Interface):
"""An object that manages a sequence of transactions.
Applications use transaction managers to establish transaction boundaries.
A transaction manager supports the "context manager" protocol:
Its `__enter__` begins a new transaction; its `__exit__` commits
the current transaction if no exception has occured; otherwise,
it aborts it.
"""
explicit = Attribute(
"""Explicit mode indicator.
This is true if the transaction manager is in explicit mode.
In explicit mode, transactions must be begun explicitly, by
calling `begin` and ended explicitly by calling
`commit` or `abort`.
.. versionadded:: 2.1.0
""")
def begin():
"""Explicitly begin and return a new transaction.
If an existing transaction is in progress and the transaction
manager not in explicit mode, the previous transaction will be
aborted. If an existing transaction is in progress and the
transaction manager is in explicit mode, an
`AlreadyInTransaction` exception will be raised..
The `~ISynchronizer.newTransaction` method of registered synchronizers
is called, passing the new transaction object.
Note that when not in explicit mode, transactions may be
started implicitly without calling `begin`. In that case,
``newTransaction`` isn't called because the transaction
manager doesn't know when to call it. The transaction is
likely to have begun long before the transaction manager is
involved. (Conceivably the `commit` and `abort` methods
could call `begin`, but they don't.)
"""
def get():
"""Get the current transaction.
In explicit mode, if a transaction hasn't begun, a
`NoTransaction` exception will be raised.
"""
def commit():
"""Commit the current transaction.
In explicit mode, if a transaction hasn't begun, a
`NoTransaction` exception will be raised.
"""
def abort():
"""Abort the current transaction.
In explicit mode, if a transaction hasn't begun, a
`NoTransaction` exception will be raised.
"""
def doom():
"""Doom the current transaction.
In explicit mode, if a transaction hasn't begun, a
`NoTransaction` exception will be raised.
"""
def isDoomed():
"""Return True if the current transaction is doomed, otherwise False.
In explicit mode, if a transaction hasn't begun, a
`NoTransaction` exception will be raised.
"""
def savepoint(optimistic=False):
"""Create a savepoint from the current transaction.
If the optimistic argument is true, then data managers that
don't support savepoints can be used, but an error will be
raised if the savepoint is rolled back.
An `ISavepoint` object is returned.
In explicit mode, if a transaction hasn't begun, a
`NoTransaction` exception will be raised.
"""
def registerSynch(synch):
"""Register an `ISynchronizer`.
Synchronizers are notified about some major events in a transaction's
life. See `ISynchronizer` for details.
If a synchronizer registers while there is an active
transaction, its ``newTransaction`` method will be called with the
active transaction.
"""
def unregisterSynch(synch):
"""Unregister an `ISynchronizer`.
Synchronizers are notified about some major events in a transaction's
life. See `ISynchronizer` for details.
"""
def clearSynchs():
"""Unregister all registered `ISynchronizer` objects.
This exists to support test cleanup/initialization
"""
def registeredSynchs():
"""Determine if any `ISynchronizers` are registered.
Return true if any are registered, and return False otherwise.
This exists to support test cleanup/initialization
"""
def attempts(number=3):
"""Generate up to *number* (transactional) context managers.
This method is typically used as follows::
for attempt in transaction_manager.attempts():
with attempt:
*with block*
The ``with attempt:`` starts a new transaction for the
execution of the *with block*. If the execution succeeds, the
(then current) transaction is commited and the ``for`` loop
terminates. If the execution raised an exception, then the
transaction is aborted. If the exception was some kind of
`retriable error <ITransaction.isRetryableError>` and the
maximal number of attempts is not yet reached, then a next
iteration of the ``for`` loop starts. In all other cases,
the ``for`` loop terminates with the exception.
"""
def run(func=None, tries=3):
"""Call *func()* in its own transaction; retry
in case of some kind of
`retriable error <ITransaction.isRetryableError>`.
The call is tried up to *tries* times.
The call is performed in a new transaction. After the call,
the (then current) transaction is committed (no exception) or
aborted (exception).
`run` supports the alternative signature ``run(tries=3)``. If
*func* is not given or passed as `None`, then the call to
`run` returns a function taking *func* as argument and then
calling ``run(func, tries)``.
"""
[docs]class ITransaction(Interface):
"""Object representing a running transaction."""
user = Attribute(
"""A user name associated with the transaction.
The format of the user name is defined by the application. The value
is text (unicode). Storages record the user value, as meta-data,
when a transaction commits.
A storage may impose a limit on the size of the value; behavior is
undefined if such a limit is exceeded (for example, a storage may
raise an exception, or truncate the value).
""")
description = Attribute(
"""A textual description of the transaction.
The value is text (unicode). Method `note` is the intended
way to set the value. Storages record the description, as meta-data,
when a transaction commits.
A storage may impose a limit on the size of the description; behavior
is undefined if such a limit is exceeded (for example, a storage may
raise an exception, or truncate the value).
""")
extension = Attribute(
"A dictionary containing application-defined metadata.")
def commit():
"""Finalize the transaction.
This executes the two-phase commit algorithm for all
`IDataManager` objects associated with the transaction.
"""
def abort():
"""Abort the transaction.
This is called from the application. This can only be called
before the two-phase commit protocol has been started.
"""
def doom():
"""Doom the transaction.
Dooms the current transaction. This will cause
`DoomedTransaction` to be raised on any attempt to commit the
transaction.
Otherwise the transaction will behave as if it was active.
"""
def savepoint(optimistic=False):
"""Create a savepoint.
If the *optimistic* argument is true, then data managers that don't
support savepoints can be used, but an error will be raised if the
savepoint is rolled back.
An `ISavepoint` object is returned.
"""
def join(datamanager):
"""Add a data manager to the transaction.
*datamanager* must provide the `IDataManager`
interface.
"""
def note(text):
"""Add text (unicode) to the transaction description.
This modifies the `description` attribute; see its docs for more
detail. First surrounding whitespace is stripped from *text*. If
`description` is currently an empty string, then the stripped text
becomes its value, else two newlines and the stripped text are
appended to `description`.
"""
def setExtendedInfo(name, value):
"""Add extension data to the transaction.
:param text name:
is the text (unicode) name of the extension property to set
:param value:
must be picklable and json serializable
Multiple calls may be made to set multiple extension
properties, provided the names are distinct.
Storages record the extension data, as meta-data, when a transaction
commits.
A storage may impose a limit on the size of extension data; behavior
is undefined if such a limit is exceeded (for example, a storage may
raise an exception, or remove `<name, value>` pairs).
"""
def addBeforeCommitHook(hook, args=(), kws=None):
"""Register a hook to call before the transaction is committed.
The specified hook function will be called after the
transaction's commit method has been called, but before the
commit process has been started.
:param sequence args:
Additional positional arguments to be passed to the hook.
The default is to pass no positional arguments.
:param dict kws:
Keyword arguments to pass to the hook. The default
is to pass no keyword arguments.
Multiple hooks can be registered and will be called in the
order they were registered (first registered, first called).
This method can also be called from a hook: an executing hook
can register more hooks. Applications should take care to
avoid creating infinite loops by recursively registering
hooks.
Hooks are called only for a top-level commit. A savepoint
creation does not call any hooks. If the transaction is
aborted, hooks are not called, and are discarded. Calling a
hook "consumes" its registration too: hook registrations do
not persist across transactions. If it's desired to call the
same hook on every transaction commit, then
`addBeforeCommitHook` must be called with that hook during
every transaction; in such a case consider registering a
synchronizer object via `ITransactionManager.registerSynch`
instead.
"""
def getBeforeCommitHooks():
"""Return iterable producing registered `addBeforeCommitHook` hooks.
A triple ``(hook, args, kws)`` is produced for each registered hook.
The hooks are produced in the order in which they would be invoked
by a top-level transaction commit.
"""
def addAfterCommitHook(hook, args=(), kws=None):
"""Register a hook to call after a transaction commit attempt.
The specified hook function will be called after the
transaction commit succeeds or aborts. The first argument
passed to the hook is a Boolean value, `True` if the commit
succeeded, or `False` if the commit aborted.
*args* and *kws* are interpreted as for `addBeforeCommitHook`
(with the exception that there is always one positional
argument, the commit status).
As with `addBeforeCommitHook`, multiple hooks can be
registered, savepoint creation doesn't call any hooks, and
calling a hook consumes its registration.
"""
def getAfterCommitHooks():
"""Return iterable producing the registered `addAfterCommitHook` hooks.
As with `getBeforeCommitHooks`, a triple ``(hook, args, kws)``
is produced for each registered hook. The hooks are produced
in the order in which they would be invoked by a top-level
transaction commit.
"""
def addBeforeAbortHook(hook, args=(), kws=None):
"""Register a hook to call before the transaction is aborted.
The specified hook function will be called after the
transaction's abort method has been called, but before the
abort process has been started.
*args* and *kws* are interpreted as for `addBeforeCommitHook`.
As with `addBeforeCommitHook`, multiple hooks can be
registered, savepoint creation doesn't call any hooks, and
calling a hook consumes its registration.
Abort hooks are called only for a top-level abort. If the
transaction is committed, abort hooks are not called. This is
true even if the commit fails. In this case, however, the
transaction is in the ``COMMITFAILED`` state and is virtually
unusable; therefore, a top-level abort will typically follow.
"""
def getBeforeAbortHooks():
"""Return iterable producing the registered `addBeforeAbortHook` hooks.
As with `getBeforeCommitHooks`, a triple ``(hook, args, kws)``
is produced for each registered hook. The hooks are produced
in the order in which they would be invoked by a top-level
transaction abort.
"""
def addAfterAbortHook(hook, args=(), kws=None):
"""Register a hook to call after a transaction abort.
The specified hook function will be called after the
transaction abort.
*args* and *kws* are interpreted as for `addBeforeCommitHook`.
As with `addBeforeCommitHook`, multiple hooks can be
registered, savepoint creation doesn't call any hooks, and
calling a hook consumes its registration.
As with `addBeforeAbortHook`, these hooks are called only for
a top-level abort. See that method for more.
"""
def getAfterAbortHooks():
"""Return iterable producing the registered `addAfterAbortHook` hooks.
As with `getBeforeCommitHooks`, a triple ``(hook, args, kws)``
is produced for each registered hook. The hooks are produced
in the order in which they would be invoked by a top-level
transaction abort.
"""
def set_data(ob, data):
"""Hold *data* on behalf of an object
For objects such as data managers or their subobjects that
work with multiple transactions, it's convenient to store
transaction-specific data on the transaction itself. The
transaction knows nothing about the data, but simply holds it
on behalf of the object.
The object passed should be the object that needs the data, as
opposed to a simple object like a string. (Internally, the id of
the object is used as the key.)
"""
def data(ob):
"""Retrieve data held on behalf of an object.
See `set_data`.
"""
def isRetryableError(error):
"""Determine if the error is retryable.
Returns true if any joined `IRetryDataManager` considers the
error transient *or* if the error is an instance of
`TransientError`. Such errors may occur due to concurrency
issues in the underlying storage engine.
"""
class IDataManager(Interface):
"""Objects that manage transactional storage.
These objects may manage data for other objects, or they may manage
non-object storages, such as relational databases. For example,
a `ZODB.Connection.Connection`.
Note that when some data is modified, that data's data manager should
join a transaction so that data can be committed when the user commits
the transaction.
These objects implement the two-phase commit protocol in order to allow
multiple data managers to safely participate in a single transaction.
The methods `tpc_begin`, `commit`, `tpc_vote`, and then either
`tpc_finish` or `tpc_abort` are normally called in that order when
committing a transaction.
"""
transaction_manager = Attribute(
"""The transaction manager (TM) used by this data manager.
This is a public attribute, intended for read-only use. The value
is an instance of `ITransactionManager`, typically set by the data
manager's constructor.
""")
def abort(transaction):
"""Abort a transaction and forget all changes.
Abort must be called outside of a two-phase commit.
Abort is called by the transaction manager to abort
transactions that are not yet in a two-phase commit. It may
also be called when rolling back a savepoint made before the
data manager joined the transaction.
In any case, after abort is called, the data manager is no
longer participating in the transaction. If there are new
changes, the data manager must rejoin the transaction.
"""
def tpc_begin(transaction):
"""Begin commit of a transaction, starting the two-phase commit.
*transaction* is the `ITransaction` instance associated with the
transaction being committed.
"""
def commit(transaction):
"""Commit modifications to registered objects.
Save changes to be made persistent if the transaction commits
(if `tpc_finish` is called later). If `tpc_abort` is called
later, changes must not persist.
This includes conflict detection and handling. If no conflicts
or errors occur, the data manager should be prepared to make
the changes persist when `tpc_finish` is called.
"""
def tpc_vote(transaction):
"""Verify that a data manager can commit the transaction.
This is the last chance for a data manager to vote 'no'. A
data manager votes 'no' by raising an exception.
*transaction* is the `ITransaction` instance associated with the
transaction being committed.
"""
def tpc_finish(transaction):
"""Indicate confirmation that the transaction is done.
Make all changes to objects modified by this transaction persist.
*transaction* is the `ITransaction` instance associated with the
transaction being committed.
This should never fail. If this raises an exception, the
database is not expected to maintain consistency; it's a
serious error.
"""
def tpc_abort(transaction):
"""Abort a transaction.
This is called by a transaction manager to end a two-phase commit on
the data manager. Abandon all changes to objects modified by this
transaction.
*transaction* is the `ITransaction` instance associated with the
transaction being committed.
This should never fail.
"""
def sortKey():
"""Return a key to use for ordering registered `IDataManagers`.
In order to guarantee a total ordering, keys **must** be
`strings <str>`.
Transactions use a global sort order to prevent deadlock when
committing transactions involving multiple data managers.
The data managers **must** define a `sortKey` method that
provides a global ordering across all registered data managers.
"""
# Alternate version:
# """Return a consistent sort key for this connection.
#
# This allows ordering multiple connections that use the same storage
# in a consistent manner. This is unique for the lifetime of a
# connection, which is good enough to avoid ZEO deadlocks.
# """
class ISavepointDataManager(IDataManager):
def savepoint():
"""Return a data-manager savepoint (`IDataManagerSavepoint`)."""
class IRetryDataManager(IDataManager):
def should_retry(exception):
"""Return whether a given exception instance should be retried.
A data manager can provide this method to indicate that a a
transaction that raised the given error should be retried.
This method may be called by an `ITransactionManager` when
considering whether to retry a failed transaction.
"""
class IDataManagerSavepoint(Interface):
"""Savepoint for data-manager changes for use in transaction savepoints.
Datamanager savepoints are used by, and only by, transaction
savepoints.
Note that data manager savepoints don't have any notion of, or
responsibility for, validity. It isn't the responsibility of
data-manager savepoints to prevent multiple rollbacks or rollbacks
after transaction termination. Preventing invalid savepoint
rollback is the responsibility of transaction rollbacks.
Application code should never use data-manager savepoints.
"""
def rollback():
"""Rollback any work done since the savepoint. """
class ISavepoint(Interface):
"""A transaction savepoint.
"""
def rollback():
"""Rollback any work done since the savepoint.
`InvalidSavepointRollbackError` is raised if the savepoint isn't valid.
"""
valid = Attribute(
"Boolean indicating whether the savepoint is valid")
class InvalidSavepointRollbackError(Exception):
"""Attempt to rollback an invalid savepoint.
A savepoint may be invalid because:
- The surrounding transaction has committed or aborted.
- An earlier savepoint in the same transaction has been rolled back.
"""
class ISynchronizer(Interface):
"""Objects that participate in the transaction-boundary notification API.
"""
def beforeCompletion(transaction):
"""Hook that is called by the transaction at the start of a commit."""
def afterCompletion(transaction):
"""Hook that is called by the transaction after completing a commit."""
def newTransaction(transaction):
"""Hook that is called at the start of a transaction.
This hook is called when, and only when, a transaction manager's
`~ITransactionManager.begin` method is called explicitly.
"""
class TransactionError(Exception):
"""An error occurred due to normal transaction processing."""
class TransactionFailedError(TransactionError):
"""Cannot perform an operation on a transaction that previously
failed.
An attempt was made to commit a transaction, or to join a
transaction, but this transaction previously raised an exception
during an attempt to commit it. The transaction must be explicitly
aborted by invoking `ITransaction.abort`. (If the transaction manager
is not operating in explicit mode, then `ITransactionManager.begin`
can also be used to perform an implicit abort.)
"""
class DoomedTransaction(TransactionError):
"""A commit was attempted on a transaction that was doomed."""
class TransientError(TransactionError):
"""An error has occured when performing a transaction.
It's possible that retrying the transaction will succeed.
"""
class NoTransaction(TransactionError):
"""No transaction has been defined
An application called an operation on a transaction manager that
affects an exciting transaction, but no transaction was begun.
The transaction manager was in explicit mode, so a new transaction
was not explicitly created.
.. versionadded:: 2.1.0
"""
class AlreadyInTransaction(TransactionError):
"""Attempt to create a new transaction without ending a preceding one
An application called `~ITransactionManager.begin` on a transaction manager
in explicit mode, without committing or aborting the previous
transaction.
.. versionadded:: 2.1.0
"""