def testDeepCopyCanInvalidate(self):
"""
Tests regression for invalidation problems related to missing
readers and writers values in cloned objects (see
http://mail.zope.org/pipermail/zodb-dev/2008-August/012054.html)
"""
import ZODB.MappingStorage
database = DB(ZODB.blob.BlobStorage(
'blobs', ZODB.MappingStorage.MappingStorage()))
connection = database.open()
root = connection.root()
transaction.begin()
root['blob'] = Blob()
transaction.commit()
stream = BytesIO()
p = Pickler(stream, _protocol)
p.dump(root['blob'])
u = Unpickler(stream)
stream.seek(0)
clone = u.load()
clone._p_invalidate()
# it should also be possible to open the cloned blob
# (even though it won't contain the original data)
clone.open().close()
# tearDown
database.close()
def cloneByPickle(obj, ignore_list=()):
"""Makes a copy of a ZODB object, loading ghosts as needed.
Ignores specified objects along the way, replacing them with None
in the copy.
"""
ignore_dict = {}
for o in ignore_list:
ignore_dict[id(o)] = o
def persistent_id(ob, ignore_dict=ignore_dict):
if id(ob) in ignore_dict:
return 'ignored'
if getattr(ob, '_p_changed', 0) is None:
ob._p_changed = 0
return None
def persistent_load(ref):
assert ref == 'ignored'
# Return a placeholder object that will be replaced by
# removeNonVersionedData().
placeholder = SimpleItem()
placeholder.id = "ignored_subobject"
return placeholder
stream = BytesIO()
p = Pickler(stream, 1)
p.persistent_id = persistent_id
p.dump(obj)
stream.seek(0)
u = Unpickler(stream)
u.persistent_load = persistent_load
return u.load()
def save(self, pos, fname):
with open(fname, 'wb') as f:
pickler = Pickler(f, _protocol)
pickler.fast = True
pickler.dump(pos)
for k, v in six.iteritems(self._data):
pickler.dump((k, v.toString()))
pickler.dump(None)
def __init__(self, obj=None):
self._file = BytesIO()
self._p = Pickler(self._file, _protocol)
if sys.version_info[0] < 3:
self._p.inst_persistent_id = self.persistent_id
else:
self._p.persistent_id = self.persistent_id
self._stack = []
if obj is not None:
self._stack.append(obj)
jar = obj._p_jar
assert myhasattr(jar, "new_oid")
self._jar = jar
def _initroot(self):
try:
self._storage.load(ZERO, '')
except KeyError:
from transaction import Transaction
file = BytesIO()
p = Pickler(file, _protocol)
p.dump((PersistentMapping, None))
p.dump({'_container': {}})
t = Transaction()
t.description = 'initial database creation'
self._storage.tpc_begin(t)
self._storage.store(ZERO, None, file.getvalue(), '', t)
self._storage.tpc_vote(t)
self._storage.tpc_finish(t)
def dumps(obj):
def getpersid(obj):
if hasattr(obj, 'getoid'):
return obj.getoid()
return None
s = BytesIO()
p = Pickler(s, _protocol)
if sys.version_info[0] < 3:
p.inst_persistent_id = getpersid
else:
p.persistent_id = getpersid
p.dump(obj)
p.dump(None)
return s.getvalue()
def __init__(self, obj=None):
self._file = BytesIO()
self._p = Pickler(self._file, _protocol)
if sys.version_info[0] < 3:
self._p.inst_persistent_id = self.persistent_id
# PyPy uses a python implementation of cPickle in both Python 2
# and Python 3. We can't really detect inst_persistent_id as its
# a magic attribute that's not readable, but it doesn't hurt to
# simply always assign to persistent_id also
self._p.persistent_id = self.persistent_id
else:
self._p.persistent_id = self.persistent_id
self._stack = []
if obj is not None:
self._stack.append(obj)
jar = obj._p_jar
assert myhasattr(jar, "new_oid")
self._jar = jar
def zodb_pickle(obj):
"""Create a pickle in the format expected by ZODB."""
f = BytesIO()
p = Pickler(f, _protocol)
if sys.version_info[0] < 3:
p.inst_persistent_id = _persistent_id
else:
p.persistent_id = _persistent_id
klass = obj.__class__
assert not hasattr(obj, '__getinitargs__'), "not ready for constructors"
args = None
mod = getattr(klass, '__module__', None)
if mod is not None:
klass = mod, klass.__name__
state = obj.__getstate__()
p.dump((klass, args))
p.dump(state)
return f.getvalue()
def make_pickle(ob):
sio = BytesIO()
p = Pickler(sio, _protocol)
p.dump(ob)
return sio.getvalue()
def pdumps(obj):
s = BytesIO()
p = Pickler(s, _protocol)
p.dump(obj)
p.dump(None)
return s.getvalue()
def tryToResolveConflict(self, oid, committedSerial, oldSerial, newpickle,
committedData=b''):
# class_tuple, old, committed, newstate = ('',''), 0, 0, 0
try:
prfactory = PersistentReferenceFactory()
newpickle = self._crs_untransform_record_data(newpickle)
file = BytesIO(newpickle)
unpickler = Unpickler(file)
unpickler.find_global = find_global
unpickler.persistent_load = prfactory.persistent_load
meta = unpickler.load()
if isinstance(meta, tuple):
klass = meta[0]
newargs = meta[1] or ()
if isinstance(klass, tuple):
klass = find_global(*klass)
else:
klass = meta
newargs = ()
if klass in _unresolvable:
raise ConflictError
inst = klass.__new__(klass, *newargs)
try:
resolve = inst._p_resolveConflict
except AttributeError:
_unresolvable[klass] = 1
raise ConflictError
oldData = self.loadSerial(oid, oldSerial)
if not committedData:
committedData = self.loadSerial(oid, committedSerial)
if newpickle == oldData:
# old -> new diff is empty, so merge is trivial
return committedData
if committedData == oldData:
# old -> committed diff is empty, so merge is trivial
return newpickle
newstate = unpickler.load()
old = state(self, oid, oldSerial, prfactory, oldData)
committed = state(self, oid, committedSerial, prfactory, committedData)
resolved = resolve(old, committed, newstate)
file = BytesIO()
pickler = Pickler(file, _protocol)
if sys.version_info[0] < 3:
pickler.inst_persistent_id = persistent_id
else:
pickler.persistent_id = persistent_id
pickler.dump(meta)
pickler.dump(resolved)
return self._crs_transform_record_data(file.getvalue())
except (ConflictError, BadClassName):
pass
except:
# If anything else went wrong, catch it here and avoid passing an
# arbitrary exception back to the client. The error here will mask
# the original ConflictError. A client can recover from a
# ConflictError, but not necessarily from other errors. But log
# the error so that any problems can be fixed.
logger.error("Unexpected error", exc_info=True)
raise ConflictError(oid=oid, serials=(committedSerial, oldSerial),
data=newpickle)
def _importDuringCommit(self, transaction, f, return_oid_list):
"""Import data during two-phase commit.
Invoked by the transaction manager mid commit.
Appends one item, the OID of the first object created,
to return_oid_list.
"""
oids = {}
# IMPORTANT: This code should be consistent with the code in
# serialize.py. It is currently out of date and doesn't handle
# weak references.
def persistent_load(ooid):
"""Remap a persistent id to a new ID and create a ghost for it."""
klass = None
if isinstance(ooid, tuple):
ooid, klass = ooid
if not isinstance(ooid, bytes):
assert isinstance(ooid, str)
# this happens on Python 3 when all bytes in the oid are < 0x80
ooid = ooid.encode('ascii')
if ooid in oids:
oid = oids[ooid]
else:
if klass is None:
oid = self._storage.new_oid()
else:
oid = self._storage.new_oid(), klass
oids[ooid] = oid
return Ghost(oid)
while 1:
header = f.read(16)
if header == export_end_marker:
break
if len(header) != 16:
raise ExportError("Truncated export file")
# Extract header information
ooid = header[:8]
length = u64(header[8:16])
data = f.read(length)
if len(data) != length:
raise ExportError("Truncated export file")
if oids:
oid = oids[ooid]
if isinstance(oid, tuple):
oid = oid[0]
else:
oids[ooid] = oid = self._storage.new_oid()
return_oid_list.append(oid)
# Blob support
blob_begin = f.read(len(blob_begin_marker))
if blob_begin == blob_begin_marker:
# Copy the blob data to a temporary file
# and remember the name
blob_len = u64(f.read(8))
blob_filename = mktemp()
blob_file = open(blob_filename, "wb")
cp(f, blob_file, blob_len)
blob_file.close()
else:
f.seek(-len(blob_begin_marker), 1)
blob_filename = None
pfile = BytesIO(data)
unpickler = Unpickler(pfile)
unpickler.persistent_load = persistent_load
newp = BytesIO()
pickler = Pickler(newp, _protocol)
if sys.version_info[0] < 3:
pickler.inst_persistent_id = persistent_id
else:
pickler.persistent_id = persistent_id
pickler.dump(unpickler.load())
pickler.dump(unpickler.load())
data = newp.getvalue()
if blob_filename is not None:
self._storage.storeBlob(oid, None, data, blob_filename, '',
transaction)
else:
self._storage.store(oid, None, data, '', transaction)
def __init__(self,
storage,
pool_size=7,
pool_timeout=1 << 31,
cache_size=400,
cache_size_bytes=0,
historical_pool_size=3,
historical_cache_size=1000,
historical_cache_size_bytes=0,
historical_timeout=300,
database_name='unnamed',
databases=None,
xrefs=True,
large_record_size=1 << 24,
**storage_args):
"""Create an object database.
:Parameters:
- `storage`: the storage used by the database, e.g. FileStorage
- `pool_size`: expected maximum number of open connections
- `cache_size`: target size of Connection object cache
- `cache_size_bytes`: target size measured in total estimated size
of objects in the Connection object cache.
"0" means unlimited.
- `historical_pool_size`: expected maximum number of total
historical connections
- `historical_cache_size`: target size of Connection object cache for
historical (`at` or `before`) connections
- `historical_cache_size_bytes` -- similar to `cache_size_bytes` for
the historical connection.
- `historical_timeout`: minimum number of seconds that
an unused historical connection will be kept, or None.
- `xrefs` - Boolian flag indicating whether implicit cross-database
references are allowed
"""
if isinstance(storage, six.string_types):
from ZODB import FileStorage
storage = ZODB.FileStorage.FileStorage(storage, **storage_args)
elif storage is None:
from ZODB import MappingStorage
storage = ZODB.MappingStorage.MappingStorage(**storage_args)
# Allocate lock.
x = threading.RLock()
self._a = x.acquire
self._r = x.release
# pools and cache sizes
self.pool = ConnectionPool(pool_size, pool_timeout)
self.historical_pool = KeyedConnectionPool(historical_pool_size,
historical_timeout)
self._cache_size = cache_size
self._cache_size_bytes = cache_size_bytes
self._historical_cache_size = historical_cache_size
self._historical_cache_size_bytes = historical_cache_size_bytes
# Setup storage
self.storage = storage
self.references = ZODB.serialize.referencesf
try:
storage.registerDB(self)
except TypeError:
storage.registerDB(self, None) # Backward compat
if (not hasattr(storage, 'tpc_vote')) and not storage.isReadOnly():
warnings.warn(
"Storage doesn't have a tpc_vote and this violates "
"the storage API. Violently monkeypatching in a do-nothing "
"tpc_vote.", DeprecationWarning, 2)
storage.tpc_vote = lambda *args: None
if IMVCCStorage.providedBy(storage):
temp_storage = storage.new_instance()
else:
temp_storage = storage
try:
try:
temp_storage.load(z64, '')
except KeyError:
# Create the database's root in the storage if it doesn't exist
from persistent.mapping import PersistentMapping
root = PersistentMapping()
# Manually create a pickle for the root to put in the storage.
# The pickle must be in the special ZODB format.
file = BytesIO()
p = Pickler(file, _protocol)
p.dump((root.__class__, None))
p.dump(root.__getstate__())
t = transaction.Transaction()
t.description = 'initial database creation'
temp_storage.tpc_begin(t)
temp_storage.store(z64, None, file.getvalue(), '', t)
temp_storage.tpc_vote(t)
temp_storage.tpc_finish(t)
finally:
if IMVCCStorage.providedBy(temp_storage):
temp_storage.release()
# Multi-database setup.
if databases is None:
databases = {}
self.databases = databases
self.database_name = database_name
if database_name in databases:
raise ValueError("database_name %r already in databases" %
database_name)
databases[database_name] = self
self.xrefs = xrefs
self.large_record_size = large_record_size
