Update of /cvsroot/python/python/dist/src/Lib/bsddb
In directory sc8-pr-cvs1:/tmp/cvs-serv24766/Lib/bsddb
Added Files:
__init__.py db.py dbobj.py dbrecio.py dbshelve.py dbtables.py
dbutils.py
Log Message:
Import PyBSDDB 3.4.0. Rename historical wrapper to bsddb185.
--- NEW FILE: __init__.py ---
#----------------------------------------------------------------------
# Copyright (c) 1999-2001, Digital Creations, Fredericksburg, VA, USA
# and Andrew Kuchling. All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are
# met:
#
# o Redistributions of source code must retain the above copyright
# notice, this list of conditions, and the disclaimer that follows.
#
# o Redistributions in binary form must reproduce the above copyright
# notice, this list of conditions, and the following disclaimer in
# the documentation and/or other materials provided with the
# distribution.
#
# o Neither the name of Digital Creations nor the names of its
# contributors may be used to endorse or promote products derived
# from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY DIGITAL CREATIONS AND CONTRIBUTORS *AS
# IS* AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
# TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
# PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL DIGITAL
# CREATIONS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
# INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
# BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS
# OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
# ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR
# TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
# USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH
# DAMAGE.
#----------------------------------------------------------------------
"""
This package initialization module provides a compatibility interface
that should enable bsddb3 to be a near drop-in replacement for the original
old bsddb module. The functions and classes provided here are all
wrappers around the new functionality provided in the bsddb3.db module.
People interested in the more advanced capabilites of Berkeley DB 3.x
should use the bsddb3.db module directly.
"""
import _bsddb
# bsddb3 calls it _db
_db = _bsddb
__version__ = _db.__version__
error = _db.DBError # So bsddb3.error will mean something...
#----------------------------------------------------------------------
class _DBWithCursor:
"""
A simple wrapper around DB that makes it look like the bsddbobject in
the old module. It uses a cursor as needed to provide DB traversal.
"""
def __init__(self, db):
self.db = db
self.dbc = None
self.db.set_get_returns_none(0)
def __del__(self):
self.close()
def _checkCursor(self):
if self.dbc is None:
self.dbc = self.db.cursor()
def _checkOpen(self):
if self.db is None:
raise error, "BSDDB object has already been closed"
def isOpen(self):
return self.db is not None
def __len__(self):
self._checkOpen()
return len(self.db)
def __getitem__(self, key):
self._checkOpen()
return self.db[key]
def __setitem__(self, key, value):
self._checkOpen()
self.db[key] = value
def __delitem__(self, key):
self._checkOpen()
del self.db[key]
def close(self):
if self.dbc is not None:
self.dbc.close()
v = 0
if self.db is not None:
v = self.db.close()
self.dbc = None
self.db = None
return v
def keys(self):
self._checkOpen()
return self.db.keys()
def has_key(self, key):
self._checkOpen()
return self.db.has_key(key)
def set_location(self, key):
self._checkOpen()
self._checkCursor()
return self.dbc.set(key)
def next(self):
self._checkOpen()
self._checkCursor()
rv = self.dbc.next()
return rv
def previous(self):
self._checkOpen()
self._checkCursor()
rv = self.dbc.prev()
return rv
def first(self):
self._checkOpen()
self._checkCursor()
rv = self.dbc.first()
return rv
def last(self):
self._checkOpen()
self._checkCursor()
rv = self.dbc.last()
return rv
def sync(self):
self._checkOpen()
return self.db.sync()
#----------------------------------------------------------------------
# Compatibility object factory functions
def hashopen(file, flag='c', mode=0666, pgsize=None, ffactor=None, nelem=None,
cachesize=None, lorder=None, hflags=0):
flags = _checkflag(flag)
d = _db.DB()
d.set_flags(hflags)
if cachesize is not None: d.set_cachesize(0, cachesize)
if pgsize is not None: d.set_pagesize(pgsize)
if lorder is not None: d.set_lorder(lorder)
if ffactor is not None: d.set_h_ffactor(ffactor)
if nelem is not None: d.set_h_nelem(nelem)
d.open(file, _db.DB_HASH, flags, mode)
return _DBWithCursor(d)
#----------------------------------------------------------------------
def btopen(file, flag='c', mode=0666,
btflags=0, cachesize=None, maxkeypage=None, minkeypage=None,
pgsize=None, lorder=None):
flags = _checkflag(flag)
d = _db.DB()
if cachesize is not None: d.set_cachesize(0, cachesize)
if pgsize is not None: d.set_pagesize(pgsize)
if lorder is not None: d.set_lorder(lorder)
d.set_flags(btflags)
if minkeypage is not None: d.set_bt_minkey(minkeypage)
if maxkeypage is not None: d.set_bt_maxkey(maxkeypage)
d.open(file, _db.DB_BTREE, flags, mode)
return _DBWithCursor(d)
#----------------------------------------------------------------------
def rnopen(file, flag='c', mode=0666,
rnflags=0, cachesize=None, pgsize=None, lorder=None,
rlen=None, delim=None, source=None, pad=None):
flags = _checkflag(flag)
d = _db.DB()
if cachesize is not None: d.set_cachesize(0, cachesize)
if pgsize is not None: d.set_pagesize(pgsize)
if lorder is not None: d.set_lorder(lorder)
d.set_flags(rnflags)
if delim is not None: d.set_re_delim(delim)
if rlen is not None: d.set_re_len(rlen)
if source is not None: d.set_re_source(source)
if pad is not None: d.set_re_pad(pad)
d.open(file, _db.DB_RECNO, flags, mode)
return _DBWithCursor(d)
#----------------------------------------------------------------------
def _checkflag(flag):
if flag == 'r':
flags = _db.DB_RDONLY
elif flag == 'rw':
flags = 0
elif flag == 'w':
flags = _db.DB_CREATE
elif flag == 'c':
flags = _db.DB_CREATE
elif flag == 'n':
flags = _db.DB_CREATE | _db.DB_TRUNCATE
else:
raise error, "flags should be one of 'r', 'w', 'c' or 'n'"
return flags | _db.DB_THREAD
#----------------------------------------------------------------------
# This is a silly little hack that allows apps to continue to use the
# DB_THREAD flag even on systems without threads without freaking out
# BerkeleyDB.
#
# This assumes that if Python was built with thread support then
# BerkeleyDB was too.
try:
import thread
del thread
except ImportError:
_db.DB_THREAD = 0
#----------------------------------------------------------------------
--- NEW FILE: db.py ---
#----------------------------------------------------------------------
# Copyright (c) 1999-2001, Digital Creations, Fredericksburg, VA, USA
# and Andrew Kuchling. All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are
# met:
#
# o Redistributions of source code must retain the above copyright
# notice, this list of conditions, and the disclaimer that follows.
#
# o Redistributions in binary form must reproduce the above copyright
# notice, this list of conditions, and the following disclaimer in
# the documentation and/or other materials provided with the
# distribution.
#
# o Neither the name of Digital Creations nor the names of its
# contributors may be used to endorse or promote products derived
# from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY DIGITAL CREATIONS AND CONTRIBUTORS *AS
# IS* AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
# TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
# PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL DIGITAL
# CREATIONS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
# INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
# BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS
# OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
# ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR
# TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
# USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH
# DAMAGE.
#----------------------------------------------------------------------
# This module is just a placeholder for possible future expansion, in
# case we ever want to augment the stuff in _db in any way. For now
# it just simply imports everything from _db.
from _db import *
from _db import __version__
if version() < (3, 1, 0):
raise ImportError, "BerkeleyDB 3.x symbols not found. Perhaps python was statically linked with an older version?"
--- NEW FILE: dbobj.py ---
#-------------------------------------------------------------------------
# This file contains real Python object wrappers for DB and DBEnv
# C "objects" that can be usefully subclassed. The previous SWIG
# based interface allowed this thanks to SWIG's shadow classes.
# -- Gregory P. Smith
#-------------------------------------------------------------------------
#
# (C) Copyright 2001 Autonomous Zone Industries
#
# License: This is free software. You may use this software for any
# purpose including modification/redistribution, so long as
# this header remains intact and that you do not claim any
# rights of ownership or authorship of this software. This
# software has been tested, but no warranty is expressed or
# implied.
#
import db
class DBEnv:
def __init__(self, *args, **kwargs):
self._cobj = apply(db.DBEnv, args, kwargs)
def close(self, *args, **kwargs):
return apply(self._cobj.close, args, kwargs)
def open(self, *args, **kwargs):
return apply(self._cobj.open, args, kwargs)
def remove(self, *args, **kwargs):
return apply(self._cobj.remove, args, kwargs)
def set_cachesize(self, *args, **kwargs):
return apply(self._cobj.set_cachesize, args, kwargs)
def set_data_dir(self, *args, **kwargs):
return apply(self._cobj.set_data_dir, args, kwargs)
def set_flags(self, *args, **kwargs):
return apply(self._cobj.set_flags, args, kwargs)
def set_lg_bsize(self, *args, **kwargs):
return apply(self._cobj.set_lg_bsize, args, kwargs)
def set_lg_dir(self, *args, **kwargs):
return apply(self._cobj.set_lg_dir, args, kwargs)
def set_lg_max(self, *args, **kwargs):
return apply(self._cobj.set_lg_max, args, kwargs)
def set_lk_detect(self, *args, **kwargs):
return apply(self._cobj.set_lk_detect, args, kwargs)
def set_lk_max(self, *args, **kwargs):
return apply(self._cobj.set_lk_max, args, kwargs)
def set_lk_max_locks(self, *args, **kwargs):
return apply(self._cobj.set_lk_max_locks, args, kwargs)
def set_lk_max_lockers(self, *args, **kwargs):
return apply(self._cobj.set_lk_max_lockers, args, kwargs)
def set_lk_max_objects(self, *args, **kwargs):
return apply(self._cobj.set_lk_max_objects, args, kwargs)
def set_mp_mmapsize(self, *args, **kwargs):
return apply(self._cobj.set_mp_mmapsize, args, kwargs)
def set_tmp_dir(self, *args, **kwargs):
return apply(self._cobj.set_tmp_dir, args, kwargs)
def txn_begin(self, *args, **kwargs):
return apply(self._cobj.txn_begin, args, kwargs)
def txn_checkpoint(self, *args, **kwargs):
return apply(self._cobj.txn_checkpoint, args, kwargs)
def txn_stat(self, *args, **kwargs):
return apply(self._cobj.txn_stat, args, kwargs)
def set_tx_max(self, *args, **kwargs):
return apply(self._cobj.set_tx_max, args, kwargs)
def lock_detect(self, *args, **kwargs):
return apply(self._cobj.lock_detect, args, kwargs)
def lock_get(self, *args, **kwargs):
return apply(self._cobj.lock_get, args, kwargs)
def lock_id(self, *args, **kwargs):
return apply(self._cobj.lock_id, args, kwargs)
def lock_put(self, *args, **kwargs):
return apply(self._cobj.lock_put, args, kwargs)
def lock_stat(self, *args, **kwargs):
return apply(self._cobj.lock_stat, args, kwargs)
def log_archive(self, *args, **kwargs):
return apply(self._cobj.log_archive, args, kwargs)
def set_get_returns_none(self, *args, **kwargs):
return apply(self._cobj.set_get_returns_none, args, kwargs)
class DB:
def __init__(self, dbenv, *args, **kwargs):
# give it the proper DBEnv C object that its expecting
self._cobj = apply(db.DB, (dbenv._cobj,) + args, kwargs)
# TODO are there other dict methods that need to be overridden?
def __len__(self):
return len(self._cobj)
def __getitem__(self, arg):
return self._cobj[arg]
def __setitem__(self, key, value):
self._cobj[key] = value
def __delitem__(self, arg):
del self._cobj[arg]
def append(self, *args, **kwargs):
return apply(self._cobj.append, args, kwargs)
def associate(self, *args, **kwargs):
return apply(self._cobj.associate, args, kwargs)
def close(self, *args, **kwargs):
return apply(self._cobj.close, args, kwargs)
def consume(self, *args, **kwargs):
return apply(self._cobj.consume, args, kwargs)
def consume_wait(self, *args, **kwargs):
return apply(self._cobj.consume_wait, args, kwargs)
def cursor(self, *args, **kwargs):
return apply(self._cobj.cursor, args, kwargs)
def delete(self, *args, **kwargs):
return apply(self._cobj.delete, args, kwargs)
def fd(self, *args, **kwargs):
return apply(self._cobj.fd, args, kwargs)
def get(self, *args, **kwargs):
return apply(self._cobj.get, args, kwargs)
def get_both(self, *args, **kwargs):
return apply(self._cobj.get_both, args, kwargs)
def get_byteswapped(self, *args, **kwargs):
return apply(self._cobj.get_byteswapped, args, kwargs)
def get_size(self, *args, **kwargs):
return apply(self._cobj.get_size, args, kwargs)
def get_type(self, *args, **kwargs):
return apply(self._cobj.get_type, args, kwargs)
def join(self, *args, **kwargs):
return apply(self._cobj.join, args, kwargs)
def key_range(self, *args, **kwargs):
return apply(self._cobj.key_range, args, kwargs)
def has_key(self, *args, **kwargs):
return apply(self._cobj.has_key, args, kwargs)
def items(self, *args, **kwargs):
return apply(self._cobj.items, args, kwargs)
def keys(self, *args, **kwargs):
return apply(self._cobj.keys, args, kwargs)
def open(self, *args, **kwargs):
return apply(self._cobj.open, args, kwargs)
def put(self, *args, **kwargs):
return apply(self._cobj.put, args, kwargs)
def remove(self, *args, **kwargs):
return apply(self._cobj.remove, args, kwargs)
def rename(self, *args, **kwargs):
return apply(self._cobj.rename, args, kwargs)
def set_bt_minkey(self, *args, **kwargs):
return apply(self._cobj.set_bt_minkey, args, kwargs)
def set_cachesize(self, *args, **kwargs):
return apply(self._cobj.set_cachesize, args, kwargs)
def set_flags(self, *args, **kwargs):
return apply(self._cobj.set_flags, args, kwargs)
def set_h_ffactor(self, *args, **kwargs):
return apply(self._cobj.set_h_ffactor, args, kwargs)
def set_h_nelem(self, *args, **kwargs):
return apply(self._cobj.set_h_nelem, args, kwargs)
def set_lorder(self, *args, **kwargs):
return apply(self._cobj.set_lorder, args, kwargs)
def set_pagesize(self, *args, **kwargs):
return apply(self._cobj.set_pagesize, args, kwargs)
def set_re_delim(self, *args, **kwargs):
return apply(self._cobj.set_re_delim, args, kwargs)
def set_re_len(self, *args, **kwargs):
return apply(self._cobj.set_re_len, args, kwargs)
def set_re_pad(self, *args, **kwargs):
return apply(self._cobj.set_re_pad, args, kwargs)
def set_re_source(self, *args, **kwargs):
return apply(self._cobj.set_re_source, args, kwargs)
def set_q_extentsize(self, *args, **kwargs):
return apply(self._cobj.set_q_extentsize, args, kwargs)
def stat(self, *args, **kwargs):
return apply(self._cobj.stat, args, kwargs)
def sync(self, *args, **kwargs):
return apply(self._cobj.sync, args, kwargs)
def type(self, *args, **kwargs):
return apply(self._cobj.type, args, kwargs)
def upgrade(self, *args, **kwargs):
return apply(self._cobj.upgrade, args, kwargs)
def values(self, *args, **kwargs):
return apply(self._cobj.values, args, kwargs)
def verify(self, *args, **kwargs):
return apply(self._cobj.verify, args, kwargs)
def set_get_returns_none(self, *args, **kwargs):
return apply(self._cobj.set_get_returns_none, args, kwargs)
--- NEW FILE: dbrecio.py ---
"""
File-like objects that read from or write to a bsddb3 record.
This implements (nearly) all stdio methods.
f = DBRecIO(db, key, txn=None)
f.close() # explicitly release resources held
flag = f.isatty() # always false
pos = f.tell() # get current position
f.seek(pos) # set current position
f.seek(pos, mode) # mode 0: absolute; 1: relative; 2: relative to EOF
buf = f.read() # read until EOF
buf = f.read(n) # read up to n bytes
f.truncate([size]) # truncate file at to at most size (default: current pos)
f.write(buf) # write at current position
f.writelines(list) # for line in list: f.write(line)
Notes:
- fileno() is left unimplemented so that code which uses it triggers
an exception early.
- There's a simple test set (see end of this file) - not yet updated
for DBRecIO.
- readline() is not implemented yet.
From:
Itamar Shtull-Trauring <itamar@maxnm.com>
"""
import errno
import string
class DBRecIO:
def __init__(self, db, key, txn=None):
self.db = db
self.key = key
self.txn = txn
self.len = None
self.pos = 0
self.closed = 0
self.softspace = 0
def close(self):
if not self.closed:
self.closed = 1
del self.db, self.txn
def isatty(self):
if self.closed:
raise ValueError, "I/O operation on closed file"
return 0
def seek(self, pos, mode = 0):
if self.closed:
raise ValueError, "I/O operation on closed file"
if mode == 1:
pos = pos + self.pos
elif mode == 2:
pos = pos + self.len
self.pos = max(0, pos)
def tell(self):
if self.closed:
raise ValueError, "I/O operation on closed file"
return self.pos
def read(self, n = -1):
if self.closed:
raise ValueError, "I/O operation on closed file"
if n < 0:
newpos = self.len
else:
newpos = min(self.pos+n, self.len)
dlen = newpos - self.pos
r = self.db.get(key, txn=self.txn, dlen=dlen, doff=self.pos)
self.pos = newpos
return r
__fixme = """
def readline(self, length=None):
if self.closed:
raise ValueError, "I/O operation on closed file"
if self.buflist:
self.buf = self.buf + string.joinfields(self.buflist, '')
self.buflist = []
i = string.find(self.buf, '\n', self.pos)
if i < 0:
newpos = self.len
else:
newpos = i+1
if length is not None:
if self.pos + length < newpos:
newpos = self.pos + length
r = self.buf[self.pos:newpos]
self.pos = newpos
return r
def readlines(self, sizehint = 0):
total = 0
lines = []
line = self.readline()
while line:
lines.append(line)
total += len(line)
if 0 < sizehint <= total:
break
line = self.readline()
return lines
"""
def truncate(self, size=None):
if self.closed:
raise ValueError, "I/O operation on closed file"
if size is None:
size = self.pos
elif size < 0:
raise IOError(errno.EINVAL,
"Negative size not allowed")
elif size < self.pos:
self.pos = size
self.db.put(key, "", txn=self.txn, dlen=self.len-size, doff=size)
def write(self, s):
if self.closed:
raise ValueError, "I/O operation on closed file"
if not s: return
if self.pos > self.len:
self.buflist.append('\0'*(self.pos - self.len))
self.len = self.pos
newpos = self.pos + len(s)
self.db.put(key, s, txn=self.txn, dlen=len(s), doff=self.pos)
self.pos = newpos
def writelines(self, list):
self.write(string.joinfields(list, ''))
def flush(self):
if self.closed:
raise ValueError, "I/O operation on closed file"
"""
# A little test suite
def _test():
import sys
if sys.argv[1:]:
file = sys.argv[1]
else:
file = '/etc/passwd'
lines = open(file, 'r').readlines()
text = open(file, 'r').read()
f = StringIO()
for line in lines[:-2]:
f.write(line)
f.writelines(lines[-2:])
if f.getvalue() != text:
raise RuntimeError, 'write failed'
length = f.tell()
print 'File length =', length
f.seek(len(lines[0]))
f.write(lines[1])
f.seek(0)
print 'First line =', `f.readline()`
here = f.tell()
line = f.readline()
print 'Second line =', `line`
f.seek(-len(line), 1)
line2 = f.read(len(line))
if line != line2:
raise RuntimeError, 'bad result after seek back'
f.seek(len(line2), 1)
list = f.readlines()
line = list[-1]
f.seek(f.tell() - len(line))
line2 = f.read()
if line != line2:
raise RuntimeError, 'bad result after seek back from EOF'
print 'Read', len(list), 'more lines'
print 'File length =', f.tell()
if f.tell() != length:
raise RuntimeError, 'bad length'
f.close()
if __name__ == '__main__':
_test()
"""
--- NEW FILE: dbshelve.py ---
#!/bin/env python
#------------------------------------------------------------------------
# Copyright (c) 1997-2001 by Total Control Software
# All Rights Reserved
#------------------------------------------------------------------------
#
# Module Name: dbShelve.py
#
# Description: A reimplementation of the standard shelve.py that
# forces the use of cPickle, and DB.
#
# Creation Date: 11/3/97 3:39:04PM
#
# License: This is free software. You may use this software for any
# purpose including modification/redistribution, so long as
# this header remains intact and that you do not claim any
# rights of ownership or authorship of this software. This
# software has been tested, but no warranty is expressed or
# implied.
#
# 13-Dec-2000: Updated to be used with the new bsddb3 package.
# Added DBShelfCursor class.
#
#------------------------------------------------------------------------
"""
Manage shelves of pickled objects using bsddb3 database files for the
storage.
"""
#------------------------------------------------------------------------
import cPickle
from bsddb3 import db
#------------------------------------------------------------------------
def open(filename, flags=db.DB_CREATE, mode=0660, filetype=db.DB_HASH,
dbenv=None, dbname=None):
"""
A simple factory function for compatibility with the standard
shleve.py module. It can be used like this, where key is a string
and data is a pickleable object:
from bsddb3 import dbshelve
db = dbshelve.open(filename)
db[key] = data
db.close()
"""
if type(flags) == type(''):
sflag = flags
if sflag == 'r':
flags = db.DB_RDONLY
elif sflag == 'rw':
flags = 0
elif sflag == 'w':
flags = db.DB_CREATE
elif sflag == 'c':
flags = db.DB_CREATE
elif sflag == 'n':
flags = db.DB_TRUNCATE | db.DB_CREATE
else:
raise error, "flags should be one of 'r', 'w', 'c' or 'n' or use the bsddb3.db.DB_* flags"
d = DBShelf(dbenv)
d.open(filename, dbname, filetype, flags, mode)
return d
#---------------------------------------------------------------------------
class DBShelf:
"""
A shelf to hold pickled objects, built upon a bsddb3 DB object. It
automatically pickles/unpickles data objects going to/from the DB.
"""
def __init__(self, dbenv=None):
self.db = db.DB(dbenv)
self.binary = 1
def __del__(self):
self.close()
def __getattr__(self, name):
"""Many methods we can just pass through to the DB object. (See below)"""
return getattr(self.db, name)
#-----------------------------------
# Dictionary access methods
def __len__(self):
return len(self.db)
def __getitem__(self, key):
data = self.db[key]
return cPickle.loads(data)
def __setitem__(self, key, value):
data = cPickle.dumps(value, self.binary)
self.db[key] = data
def __delitem__(self, key):
del self.db[key]
def keys(self, txn=None):
if txn != None:
return self.db.keys(txn)
else:
return self.db.keys()
def items(self, txn=None):
if txn != None:
items = self.db.items(txn)
else:
items = self.db.items()
newitems = []
for k, v in items:
newitems.append( (k, cPickle.loads(v)) )
return newitems
def values(self, txn=None):
if txn != None:
values = self.db.values(txn)
else:
values = self.db.values()
return map(cPickle.loads, values)
#-----------------------------------
# Other methods
def append(self, value, txn=None):
data = cPickle.dumps(value, self.binary)
return self.db.append(data, txn)
def associate(self, secondaryDB, callback, flags=0):
def _shelf_callback(priKey, priData, realCallback=callback):
data = cPickle.loads(priData)
return realCallback(priKey, data)
return self.db.associate(secondaryDB, _shelf_callback, flags)
#def get(self, key, default=None, txn=None, flags=0):
def get(self, *args, **kw):
# We do it with *args and **kw so if the default value wasn't
# given nothing is passed to the extension module. That way
# an exception can be raised if set_get_returns_none is turned
# off.
data = apply(self.db.get, args, kw)
try:
return cPickle.loads(data)
except (TypeError, cPickle.UnpicklingError):
return data # we may be getting the default value, or None,
# so it doesn't need unpickled.
def get_both(self, key, value, txn=None, flags=0):
data = cPickle.dumps(value, self.binary)
data = self.db.get(key, data, txn, flags)
return cPickle.loads(data)
def cursor(self, txn=None, flags=0):
c = DBShelfCursor(self.db.cursor(txn, flags))
c.binary = self.binary
return c
def put(self, key, value, txn=None, flags=0):
data = cPickle.dumps(value, self.binary)
return self.db.put(key, data, txn, flags)
def join(self, cursorList, flags=0):
raise NotImplementedError
#----------------------------------------------
# Methods allowed to pass-through to self.db
#
# close, delete, fd, get_byteswapped, get_type, has_key,
# key_range, open, remove, rename, stat, sync,
# upgrade, verify, and all set_* methods.
#---------------------------------------------------------------------------
class DBShelfCursor:
"""
"""
def __init__(self, cursor):
self.dbc = cursor
def __del__(self):
self.close()
def __getattr__(self, name):
"""Some methods we can just pass through to the cursor object. (See below)"""
return getattr(self.dbc, name)
#----------------------------------------------
def dup(self, flags=0):
return DBShelfCursor(self.dbc.dup(flags))
def put(self, key, value, flags=0):
data = cPickle.dumps(value, self.binary)
return self.dbc.put(key, data, flags)
def get(self, *args):
count = len(args) # a method overloading hack
method = getattr(self, 'get_%d' % count)
apply(method, args)
def get_1(self, flags):
rec = self.dbc.get(flags)
return self._extract(rec)
def get_2(self, key, flags):
rec = self.dbc.get(key, flags)
return self._extract(rec)
def get_3(self, key, value, flags):
data = cPickle.dumps(value, self.binary)
rec = self.dbc.get(key, flags)
return self._extract(rec)
def current(self, flags=0): return self.get_1(flags|db.DB_CURRENT)
def first(self, flags=0): return self.get_1(flags|db.DB_FIRST)
def last(self, flags=0): return self.get_1(flags|db.DB_LAST)
def next(self, flags=0): return self.get_1(flags|db.DB_NEXT)
def prev(self, flags=0): return self.get_1(flags|db.DB_PREV)
def consume(self, flags=0): return self.get_1(flags|db.DB_CONSUME)
def next_dup(self, flags=0): return self.get_1(flags|db.DB_NEXT_DUP)
def next_nodup(self, flags=0): return self.get_1(flags|db.DB_NEXT_NODUP)
def prev_nodup(self, flags=0): return self.get_1(flags|db.DB_PREV_NODUP)
def get_both(self, key, value, flags=0):
data = cPickle.dumps(value, self.binary)
rec = self.dbc.get_both(key, flags)
return self._extract(rec)
def set(self, key, flags=0):
rec = self.dbc.set(key, flags)
return self._extract(rec)
def set_range(self, key, flags=0):
rec = self.dbc.set_range(key, flags)
return self._extract(rec)
def set_recno(self, recno, flags=0):
rec = self.dbc.set_recno(recno, flags)
return self._extract(rec)
set_both = get_both
def _extract(self, rec):
if rec is None:
return None
else:
key, data = rec
return key, cPickle.loads(data)
#----------------------------------------------
# Methods allowed to pass-through to self.dbc
#
# close, count, delete, get_recno, join_item
#---------------------------------------------------------------------------
--- NEW FILE: dbtables.py ---
#-----------------------------------------------------------------------
#
# Copyright (C) 2000, 2001 by Autonomous Zone Industries
#
# License: This is free software. You may use this software for any
# purpose including modification/redistribution, so long as
# this header remains intact and that you do not claim any
# rights of ownership or authorship of this software. This
# software has been tested, but no warranty is expressed or
# implied.
#
# -- Gregory P. Smith <greg@electricrain.com>
# This provides a simple database table interface built on top of
# the Python BerkeleyDB 3 interface.
#
_cvsid = '$Id: dbtables.py,v 1.1 2002/11/19 08:09:52 loewis Exp $'
import string
import sys
try:
import cPickle
pickle = cPickle
except ImportError:
import pickle
import whrandom
import xdrlib
import re
import copy
from bsddb3.db import *
class TableDBError(StandardError): pass
class TableAlreadyExists(TableDBError): pass
class Cond:
"""This condition matches everything"""
def __call__(self, s):
return 1
class ExactCond(Cond):
"""Acts as an exact match condition function"""
def __init__(self, strtomatch):
self.strtomatch = strtomatch
def __call__(self, s):
return s == self.strtomatch
class PrefixCond(Cond):
"""Acts as a condition function for matching a string prefix"""
def __init__(self, prefix):
self.prefix = prefix
def __call__(self, s):
return s[:len(self.prefix)] == self.prefix
class LikeCond(Cond):
"""
Acts as a function that will match using an SQL 'LIKE' style
string. Case insensitive and % signs are wild cards.
This isn't perfect but it should work for the simple common cases.
"""
def __init__(self, likestr, re_flags=re.IGNORECASE):
# escape python re characters
chars_to_escape = '.*+()[]?'
for char in chars_to_escape :
likestr = string.replace(likestr, char, '\\'+char)
# convert %s to wildcards
self.likestr = string.replace(likestr, '%', '.*')
self.re = re.compile('^'+self.likestr+'$', re_flags)
def __call__(self, s):
return self.re.match(s)
#
# keys used to store database metadata
#
_table_names_key = '__TABLE_NAMES__' # list of the tables in this db
_columns = '._COLUMNS__' # table_name+this key contains a list of columns
def _columns_key(table) : return table + _columns
#
# these keys are found within table sub databases
#
_data = '._DATA_.' # this+column+this+rowid key contains table data
_rowid = '._ROWID_.' # this+rowid+this key contains a unique entry for each
# row in the table. (no data is stored)
_rowid_str_len = 8 # length in bytes of the unique rowid strings
def _data_key(table, col, rowid) : return table + _data + col + _data + rowid
def _search_col_data_key(table, col) : return table + _data + col + _data
def _search_all_data_key(table) : return table + _data
def _rowid_key(table, rowid) : return table + _rowid + rowid + _rowid
def _search_rowid_key(table) : return table + _rowid
def contains_metastrings(s) :
"""Verify that the given string does not contain any
metadata strings that might interfere with dbtables database operation.
"""
if string.find(s, _table_names_key) >= 0 or \
string.find(s, _columns) >= 0 or \
string.find(s, _data) >= 0 or \
string.find(s, _rowid) >= 0 :
return 1
else :
return 0
class bsdTableDB :
def __init__(self, filename, dbhome, create=0, truncate=0, mode=0600, recover=0, dbflags=0) :
"""bsdTableDB.open(filename, dbhome, create=0, truncate=0, mode=0600)
Open database name in the dbhome BerkeleyDB directory.
Use keyword arguments when calling this constructor.
"""
myflags = DB_THREAD
if create :
myflags = myflags | DB_CREATE
flagsforenv = DB_INIT_MPOOL | DB_INIT_LOCK | DB_INIT_LOG | DB_INIT_TXN | dbflags
if recover :
flagsforenv = flagsforenv | DB_RECOVER
self.env = DBEnv()
self.env.set_lk_detect(DB_LOCK_DEFAULT) # enable auto deadlock avoidance
self.env.open(dbhome, myflags | flagsforenv)
if truncate :
myflags = myflags | DB_TRUNCATE
self.db = DB(self.env)
self.db.set_flags(DB_DUP) # allow duplicate entries [warning: be careful w/ metadata]
self.db.open(filename, DB_BTREE, myflags, mode)
self.dbfilename = filename
# Initialize the table names list if this is a new database
if not self.db.has_key(_table_names_key) :
self.db.put(_table_names_key, pickle.dumps([], 1))
# TODO verify more of the database's metadata?
self.__tablecolumns = {}
def __del__(self):
self.close()
def close(self):
if self.db is not None:
self.db.close()
self.db = None
if self.env is not None:
self.env.close()
self.env = None
def checkpoint(self, mins=0):
try:
self.env.txn_checkpoint(mins)
except DBIncompleteError:
pass
def sync(self):
try:
self.db.sync()
except DBIncompleteError:
pass
def _db_print(self) :
"""Print the database to stdout for debugging"""
print "******** Printing raw database for debugging ********"
cur = self.db.cursor()
try:
key, data = cur.first()
while 1 :
print `{key: data}`
next = cur.next()
if next:
key, data = next
else:
cur.close()
return
except DBNotFoundError:
cur.close()
def CreateTable(self, table, columns) :
"""CreateTable(table, columns) - Create a new table in the database
raises TableDBError if it already exists or for other DB errors.
"""
assert type(columns) == type([])
txn = None
try:
# checking sanity of the table and column names here on
# table creation will prevent problems elsewhere.
if contains_metastrings(table) :
raise ValueError, "bad table name: contains reserved metastrings"
for column in columns :
if contains_metastrings(column) :
raise ValueError, "bad column name: contains reserved metastrings"
columnlist_key = _columns_key(table)
if self.db.has_key(columnlist_key) :
raise TableAlreadyExists, "table already exists"
txn = self.env.txn_begin()
# store the table's column info
self.db.put(columnlist_key, pickle.dumps(columns, 1), txn=txn)
# add the table name to the tablelist
tablelist = pickle.loads(self.db.get(_table_names_key, txn=txn, flags=DB_RMW))
tablelist.append(table)
self.db.delete(_table_names_key, txn) # delete 1st, incase we opened with DB_DUP
self.db.put(_table_names_key, pickle.dumps(tablelist, 1), txn=txn)
txn.commit()
txn = None
except DBError, dberror:
if txn :
txn.abort()
raise TableDBError, dberror[1]
def ListTableColumns(self, table):
"""Return a list of columns in the given table. [] if the table doesn't exist.
"""
assert type(table) == type('')
if contains_metastrings(table) :
raise ValueError, "bad table name: contains reserved metastrings"
columnlist_key = _columns_key(table)
if not self.db.has_key(columnlist_key):
return []
pickledcolumnlist = self.db.get(columnlist_key)
if pickledcolumnlist:
return pickle.loads(pickledcolumnlist)
else:
return []
def ListTables(self):
"""Return a list of tables in this database."""
pickledtablelist = self.db.get(_table_names_key)
if pickledtablelist:
return pickle.loads(pickledtablelist)
else:
return []
def CreateOrExtendTable(self, table, columns):
"""CreateOrExtendTable(table, columns) - Create a new table in the database.
If a table of this name already exists, extend it to have any
additional columns present in the given list as well as
all of its current columns.
"""
assert type(columns) == type([])
try:
self.CreateTable(table, columns)
except TableAlreadyExists:
# the table already existed, add any new columns
txn = None
try:
columnlist_key = _columns_key(table)
txn = self.env.txn_begin()
# load the current column list
oldcolumnlist = pickle.loads(self.db.get(columnlist_key, txn=txn, flags=DB_RMW))
# create a hash table for fast lookups of column names in the loop below
oldcolumnhash = {}
for c in oldcolumnlist:
oldcolumnhash[c] = c
# create a new column list containing both the old and new column names
newcolumnlist = copy.copy(oldcolumnlist)
for c in columns:
if not oldcolumnhash.has_key(c):
newcolumnlist.append(c)
# store the table's new extended column list
if newcolumnlist != oldcolumnlist :
# delete the old one first since we opened with DB_DUP
self.db.delete(columnlist_key, txn)
self.db.put(columnlist_key, pickle.dumps(newcolumnlist, 1), txn=txn)
txn.commit()
txn = None
self.__load_column_info(table)
except DBError, dberror:
if txn:
txn.abort()
raise TableDBError, dberror[1]
def __load_column_info(self, table) :
"""initialize the self.__tablecolumns dict"""
# check the column names
try:
tcolpickles = self.db.get(_columns_key(table))
except DBNotFoundError:
raise TableDBError, "unknown table: " + `table`
if not tcolpickles:
raise TableDBError, "unknown table: " + `table`
self.__tablecolumns[table] = pickle.loads(tcolpickles)
def __new_rowid(self, table, txn=None) :
"""Create a new unique row identifier"""
unique = 0
while not unique :
# Generate a random 64-bit row ID string
# (note: this code has <64 bits of randomness
# but it's plenty for our database id needs!)
p = xdrlib.Packer()
p.pack_int(int(whrandom.random()*2147483647))
p.pack_int(int(whrandom.random()*2147483647))
newid = p.get_buffer()
# Guarantee uniqueness by adding this key to the database
try:
self.db.put(_rowid_key(table, newid), None, txn=txn, flags=DB_NOOVERWRITE)
except DBKeyExistsError:
pass
else:
unique = 1
return newid
def Insert(self, table, rowdict) :
"""Insert(table, datadict) - Insert a new row into the table
using the keys+values from rowdict as the column values.
"""
txn = None
try:
if not self.db.has_key(_columns_key(table)) :
raise TableDBError, "unknown table"
# check the validity of each column name
if not self.__tablecolumns.has_key(table) :
self.__load_column_info(table)
for column in rowdict.keys() :
if not self.__tablecolumns[table].count(column) :
raise TableDBError, "unknown column: "+`column`
# get a unique row identifier for this row
rowid = self.__new_rowid(table)
txn = self.env.txn_begin()
# insert the row values into the table database
for column, dataitem in rowdict.items() :
# store the value
self.db.put(_data_key(table, column, rowid), dataitem, txn=txn)
txn.commit()
txn = None
except DBError, dberror:
if txn :
txn.abort()
self.db.delete(_rowid_key(table, rowid))
raise TableDBError, dberror[1]
def Modify(self, table, conditions={}, mappings={}) :
"""Modify(table, conditions) - Modify in rows matching 'conditions'
using mapping functions in 'mappings'
* conditions is a dictionary keyed on column names
containing condition functions expecting the data string as an
argument and returning a boolean.
* mappings is a dictionary keyed on column names containint condition
functions expecting the data string as an argument and returning the
new string for that column.
"""
try:
matching_rowids = self.__Select(table, [], conditions)
# modify only requested columns
columns = mappings.keys()
for rowid in matching_rowids.keys() :
txn = None
try:
for column in columns :
txn = self.env.txn_begin()
# modify the requested column
try:
dataitem = self.db.get(_data_key(table, column, rowid), txn)
self.db.delete(_data_key(table, column, rowid), txn)
except DBNotFoundError:
dataitem = None # XXXXXXX row key somehow didn't exist, assume no error
dataitem = mappings[column](dataitem)
if dataitem <> None:
self.db.put(_data_key(table, column, rowid), dataitem, txn=txn)
txn.commit()
txn = None
except DBError, dberror:
if txn :
txn.abort()
raise
except DBError, dberror:
raise TableDBError, dberror[1]
def Delete(self, table, conditions={}) :
"""Delete(table, conditions) - Delete items matching the given
conditions from the table.
* conditions is a dictionary keyed on column names
containing condition functions expecting the data string as an
argument and returning a boolean.
"""
try:
matching_rowids = self.__Select(table, [], conditions)
# delete row data from all columns
columns = self.__tablecolumns[table]
for rowid in matching_rowids.keys() :
txn = None
try:
txn = self.env.txn_begin()
for column in columns :
# delete the data key
try:
self.db.delete(_data_key(table, column, rowid), txn)
except DBNotFoundError:
pass # XXXXXXX column may not exist, assume no error
try:
self.db.delete(_rowid_key(table, rowid), txn)
except DBNotFoundError:
pass # XXXXXXX row key somehow didn't exist, assume no error
txn.commit()
txn = None
except DBError, dberror:
if txn :
txn.abort()
raise
except DBError, dberror:
raise TableDBError, dberror[1]
def Select(self, table, columns, conditions={}) :
"""Select(table, conditions) - retrieve specific row data
Returns a list of row column->value mapping dictionaries.
* columns is a list of which column data to return. If
columns is None, all columns will be returned.
* conditions is a dictionary keyed on column names
containing callable conditions expecting the data string as an
argument and returning a boolean.
"""
try:
if not self.__tablecolumns.has_key(table) :
self.__load_column_info(table)
if columns is None :
columns = self.__tablecolumns[table]
matching_rowids = self.__Select(table, columns, conditions)
except DBError, dberror:
raise TableDBError, dberror[1]
# return the matches as a list of dictionaries
return matching_rowids.values()
def __Select(self, table, columns, conditions) :
"""__Select() - Used to implement Select and Delete (above)
Returns a dictionary keyed on rowids containing dicts
holding the row data for columns listed in the columns param
that match the given conditions.
* conditions is a dictionary keyed on column names
containing callable conditions expecting the data string as an
argument and returning a boolean.
"""
# check the validity of each column name
if not self.__tablecolumns.has_key(table) :
self.__load_column_info(table)
if columns is None :
columns = self.tablecolumns[table]
for column in (columns + conditions.keys()) :
if not self.__tablecolumns[table].count(column) :
raise TableDBError, "unknown column: "+`column`
# keyed on rows that match so far, containings dicts keyed on
# column names containing the data for that row and column.
matching_rowids = {}
rejected_rowids = {} # keys are rowids that do not match
# attempt to sort the conditions in such a way as to minimize full column lookups
def cmp_conditions(atuple, btuple):
a = atuple[1]
b = btuple[1]
if type(a) == type(b) :
if isinstance(a, PrefixCond) and isinstance(b, PrefixCond):
return cmp(len(b.prefix), len(a.prefix)) # longest prefix first
if isinstance(a, LikeCond) and isinstance(b, LikeCond):
return cmp(len(b.likestr), len(a.likestr)) # longest likestr first
return 0
if isinstance(a, ExactCond):
return -1
if isinstance(b, ExactCond):
return 1
if isinstance(a, PrefixCond):
return -1
if isinstance(b, PrefixCond):
return 1
# leave all unknown condition callables alone as equals
return 0
conditionlist = conditions.items()
conditionlist.sort(cmp_conditions)
# Apply conditions to column data to find what we want
cur = self.db.cursor()
column_num = -1
for column, condition in conditionlist :
column_num = column_num + 1
searchkey = _search_col_data_key(table, column)
# speedup: don't linear search columns within loop
if column in columns :
savethiscolumndata = 1 # save the data for return
else :
savethiscolumndata = 0 # data only used for selection
try:
key, data = cur.set_range(searchkey)
while key[:len(searchkey)] == searchkey :
# extract the rowid from the key
rowid = key[-_rowid_str_len:]
if not rejected_rowids.has_key(rowid) :
# if no condition was specified or the condition
# succeeds, add row to our match list.
if not condition or condition(data) :
# only create new entries in matcing_rowids on
# the first pass, otherwise reject the
# rowid as it must not have matched
# the previous passes
if column_num == 0 :
if not matching_rowids.has_key(rowid) :
matching_rowids[rowid] = {}
if savethiscolumndata :
matching_rowids[rowid][column] = data
else :
rejected_rowids[rowid] = rowid
else :
if matching_rowids.has_key(rowid) :
del matching_rowids[rowid]
rejected_rowids[rowid] = rowid
key, data = cur.next()
except DBError, dberror:
if dberror[0] != DB_NOTFOUND :
raise
continue
cur.close()
# we're done selecting rows, garbage collect the reject list
del rejected_rowids
# extract any remaining desired column data from the
# database for the matching rows.
if len(columns) > 0 :
for rowid, rowdata in matching_rowids.items() :
for column in columns :
if rowdata.has_key(column) :
continue
try:
rowdata[column] = self.db.get(_data_key(table, column, rowid))
except DBError, dberror:
if dberror[0] != DB_NOTFOUND :
raise
rowdata[column] = None
# return the matches
return matching_rowids
def Drop(self, table) :
"""Remove an entire table from the database
"""
txn = None
try:
txn = self.env.txn_begin()
# delete the column list
self.db.delete(_columns_key(table), txn)
cur = self.db.cursor(txn)
# delete all keys containing this tables column and row info
table_key = _search_all_data_key(table)
while 1 :
try:
key, data = cur.set_range(table_key)
except DBNotFoundError:
break
# only delete items in this table
if key[:len(table_key)] != table_key :
break
cur.delete()
# delete all rowids used by this table
table_key = _search_rowid_key(table)
while 1 :
try:
key, data = cur.set_range(table_key)
except DBNotFoundError:
break
# only delete items in this table
if key[:len(table_key)] != table_key :
break
cur.delete()
cur.close()
# delete the tablename from the table name list
tablelist = pickle.loads(self.db.get(_table_names_key, txn=txn, flags=DB_RMW))
try:
tablelist.remove(table)
except ValueError:
pass # hmm, it wasn't there, oh well, that's what we want.
self.db.delete(_table_names_key, txn) # delete 1st, incase we opened with DB_DUP
self.db.put(_table_names_key, pickle.dumps(tablelist, 1), txn=txn)
txn.commit()
txn = None
if self.__tablecolumns.has_key(table) :
del self.__tablecolumns[table]
except DBError, dberror:
if txn :
txn.abort()
raise TableDBError, dberror[1]
--- NEW FILE: dbutils.py ---
#------------------------------------------------------------------------
#
# In my performance tests, using this (as in dbtest.py test4) is
# slightly slower than simply compiling _db.c with MYDB_THREAD
# undefined to prevent multithreading support in the C module.
# Using NoDeadlockDb also prevent deadlocks from mutliple processes
# accessing the same database.
#
# Copyright (C) 2000 Autonomous Zone Industries
#
# License: This is free software. You may use this software for any
# purpose including modification/redistribution, so long as
# this header remains intact and that you do not claim any
# rights of ownership or authorship of this software. This
# software has been tested, but no warranty is expressed or
# implied.
#
# Author: Gregory P. Smith <greg@electricrain.com>
#
# Note: I don't know how useful this is in reality since when a
# DBDeadlockError happens the current transaction is supposed to be
# aborted. If it doesn't then when the operation is attempted again
# the deadlock is still happening...
# --Robin
#
#------------------------------------------------------------------------
#
# import the time.sleep function in a namespace safe way to allow
# "from bsddb3.db import *"
#
from time import sleep
_sleep = sleep
del sleep
import _db
_deadlock_MinSleepTime = 1.0/64 # always sleep at least N seconds between retrys
_deadlock_MaxSleepTime = 1.0 # never sleep more than N seconds between retrys
def DeadlockWrap(function, *_args, **_kwargs):
"""DeadlockWrap(function, *_args, **_kwargs) - automatically retries
function in case of a database deadlock.
This is a DeadlockWrapper method which DB calls can be made using to
preform infinite retrys with sleeps in between when a DBLockDeadlockError
exception is raised in a database call:
d = DB(...)
d.open(...)
DeadlockWrap(d.put, "foo", data="bar") # set key "foo" to "bar"
"""
sleeptime = _deadlock_MinSleepTime
while (1) :
try:
return apply(function, _args, _kwargs)
except _db.DBLockDeadlockError:
print 'DeadlockWrap sleeping ', sleeptime
_sleep(sleeptime)
# exponential backoff in the sleep time
sleeptime = sleeptime * 2
if sleeptime > _deadlock_MaxSleepTime :
sleeptime = _deadlock_MaxSleepTime
#------------------------------------------------------------------------
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4