PY3 = sys.version_info[0] == 3
MAX_INT32 = 2147483647
MIN_INT32 = -2147483648
MAX_INT64 = 9223372036854775807
MIN_INT64 = -9223372036854775808
EPOCH_AWARE = datetime.datetime.fromtimestamp(0, utc)
EPOCH_NAIVE = datetime.datetime.utcfromtimestamp(0)
# Create constants compatible with all versions of
# python from 2.4 forward. In 2.x b("foo") is just
# "foo". In 3.x it becomes b"foo".
EMPTY = b("")
ZERO = b("\x00")
ONE = b("\x01")
BSONNUM = b("\x01") # Floating point
BSONSTR = b("\x02") # UTF-8 string
BSONOBJ = b("\x03") # Embedded document
BSONARR = b("\x04") # Array
BSONBIN = b("\x05") # Binary
BSONUND = b("\x06") # Undefined
BSONOID = b("\x07") # ObjectId
BSONBOO = b("\x08") # Boolean
BSONDAT = b("\x09") # UTC Datetime
BSONNUL = b("\x0A") # Null
BSONRGX = b("\x0B") # Regex
BSONREF = b("\x0C") # DBRef
def _element_to_bson(key, value, check_keys, uuid_subtype):
if not isinstance(key, basestring):
raise InvalidDocument("documents must have only string keys, " "key was %r" % key)
if check_keys:
if key.startswith("$"):
raise InvalidDocument("key %r must not start with '$'" % key)
if "." in key:
raise InvalidDocument("key %r must not contain '.'" % key)
name = _make_c_string(key, True)
if isinstance(value, float):
return BSONNUM + name + struct.pack("<d", value)
if _use_uuid:
if isinstance(value, uuid.UUID):
# Java Legacy
if uuid_subtype == JAVA_LEGACY:
# Python 3.0(.1) returns a bytearray instance for bytes (3.1
# and newer just return a bytes instance). Convert that to
# binary_type (here and below) for compatibility.
from_uuid = binary_type(value.bytes)
as_legacy_java = from_uuid[0:8][::-1] + from_uuid[8:16][::-1]
value = Binary(as_legacy_java, subtype=OLD_UUID_SUBTYPE)
# C# legacy
elif uuid_subtype == CSHARP_LEGACY:
# Microsoft GUID representation.
value = Binary(binary_type(value.bytes_le), subtype=OLD_UUID_SUBTYPE)
# Python
else:
value = Binary(binary_type(value.bytes), subtype=uuid_subtype)
if isinstance(value, Binary):
subtype = value.subtype
if subtype == 2:
value = struct.pack("<i", len(value)) + value
return BSONBIN + name + struct.pack("<i", len(value)) + b(chr(subtype)) + value
if isinstance(value, Code):
cstring = _make_c_string(value)
if not value.scope:
length = struct.pack("<i", len(cstring))
return BSONCOD + name + length + cstring
scope = _dict_to_bson(value.scope, False, uuid_subtype, False)
full_length = struct.pack("<i", 8 + len(cstring) + len(scope))
length = struct.pack("<i", len(cstring))
return BSONCWS + name + full_length + length + cstring + scope
if isinstance(value, binary_type):
if PY3:
# Python3 special case. Store 'bytes' as BSON binary subtype 0.
return BSONBIN + name + struct.pack("<i", len(value)) + ZERO + value
cstring = _make_c_string(value)
length = struct.pack("<i", len(cstring))
return BSONSTR + name + length + cstring
if isinstance(value, unicode):
cstring = _make_c_string(value)
length = struct.pack("<i", len(cstring))
return BSONSTR + name + length + cstring
if isinstance(value, dict):
return BSONOBJ + name + _dict_to_bson(value, check_keys, uuid_subtype, False)
if isinstance(value, (list, tuple)):
as_dict = SON(zip([str(i) for i in range(len(value))], value))
return BSONARR + name + _dict_to_bson(as_dict, check_keys, uuid_subtype, False)
if isinstance(value, ObjectId):
return BSONOID + name + value.binary
if value is True:
return BSONBOO + name + ONE
if value is False:
return BSONBOO + name + ZERO
if isinstance(value, int):
# TODO this is an ugly way to check for this...
if value > MAX_INT64 or value < MIN_INT64:
raise OverflowError("BSON can only handle up to 8-byte ints")
if value > MAX_INT32 or value < MIN_INT32:
return BSONLON + name + struct.pack("<q", value)
return BSONINT + name + struct.pack("<i", value)
# 2to3 will convert long to int here since there is no long in python3.
# That's OK. The previous if block will match instead.
if isinstance(value, long):
if value > MAX_INT64 or value < MIN_INT64:
raise OverflowError("BSON can only handle up to 8-byte ints")
return BSONLON + name + struct.pack("<q", value)
if isinstance(value, datetime.datetime):
if value.utcoffset() is not None:
value = value - value.utcoffset()
millis = int(calendar.timegm(value.timetuple()) * 1000 + value.microsecond / 1000)
return BSONDAT + name + struct.pack("<q", millis)
if isinstance(value, Timestamp):
time = struct.pack("<I", value.time)
inc = struct.pack("<I", value.inc)
return BSONTIM + name + inc + time
if value is None:
return BSONNUL + name
if isinstance(value, RE_TYPE):
pattern = value.pattern
flags = ""
if value.flags & re.IGNORECASE:
flags += "i"
if value.flags & re.LOCALE:
flags += "l"
if value.flags & re.MULTILINE:
flags += "m"
if value.flags & re.DOTALL:
flags += "s"
if value.flags & re.UNICODE:
flags += "u"
if value.flags & re.VERBOSE:
flags += "x"
return BSONRGX + name + _make_c_string(pattern, True) + _make_c_string(flags)
if isinstance(value, DBRef):
return _element_to_bson(key, value.as_doc(), False, uuid_subtype)
if isinstance(value, MinKey):
return BSONMIN + name
if isinstance(value, MaxKey):
return BSONMAX + name
raise InvalidDocument("cannot convert value of type %s to xtbson" % type(value))
except ImportError: # for Python < 2.5
import md5
_md5func = md5.new
import os
import random
import socket
import struct
import threading
import time
from xtbson.errors import InvalidId
from xtbson.py3compat import (PY3, b, binary_type, text_type,
bytes_from_hex, string_types)
from xtbson.tz_util import utc
EMPTY = b("")
ZERO = b("\x00")
def _machine_bytes():
"""Get the machine portion of an ObjectId.
"""
machine_hash = _md5func()
if PY3:
# gethostname() returns a unicode string in python 3.x
# while update() requires a byte string.
machine_hash.update(socket.gethostname().encode())
else:
# Calling encode() here will fail with non-ascii hostnames
machine_hash.update(socket.gethostname())
return machine_hash.digest()[0:3]
