from datetime import datetime as _DateTime
import sys
import struct
PY2 = sys.version_info[0] == 2
if PY2:
int_types = (int, long)
def dict_iteritems(d):
return d.iteritems()
else:
int_types = int
unicode = str
xrange = range
def dict_iteritems(d):
return d.items()
if sys.version_info < (3, 5):
RecursionError = RuntimeError
def _is_recursionerror(e):
return (
len(e.args) == 1
and isinstance(e.args[0], str)
and e.args[0].startswith("maximum recursion depth exceeded")
)
else:
def _is_recursionerror(e):
return True
if hasattr(sys, "pypy_version_info"):
from __pypy__ import newlist_hint
try:
from __pypy__.builders import BytesBuilder as StringBuilder
except ImportError:
from __pypy__.builders import StringBuilder
USING_STRINGBUILDER = True
class StringIO(object):
def __init__(self, s=b""):
if s:
self.builder = StringBuilder(len(s))
self.builder.append(s)
else:
self.builder = StringBuilder()
def write(self, s):
if isinstance(s, memoryview):
s = s.tobytes()
elif isinstance(s, bytearray):
s = bytes(s)
self.builder.append(s)
def getvalue(self):
return self.builder.build()
else:
USING_STRINGBUILDER = False
from io import BytesIO as StringIO
newlist_hint = lambda size: []
from .exceptions import BufferFull, OutOfData, ExtraData, FormatError, StackError
from .ext import ExtType, Timestamp
EX_SKIP = 0
EX_CONSTRUCT = 1
EX_READ_ARRAY_HEADER = 2
EX_READ_MAP_HEADER = 3
TYPE_IMMEDIATE = 0
TYPE_ARRAY = 1
TYPE_MAP = 2
TYPE_RAW = 3
TYPE_BIN = 4
TYPE_EXT = 5
DEFAULT_RECURSE_LIMIT = 511
def _check_type_strict(obj, t, type=type, tuple=tuple):
if type(t) is tuple:
return type(obj) in t
else:
return type(obj) is t
def _get_data_from_buffer(obj):
view = memoryview(obj)
if view.itemsize != 1:
raise ValueError("cannot unpack from multi-byte object")
return view
def unpackb(packed, **kwargs):
unpacker = Unpacker(None, max_buffer_size=len(packed), **kwargs)
unpacker.feed(packed)
try:
ret = unpacker._unpack()
except OutOfData:
raise ValueError("Unpack failed: incomplete input")
except RecursionError as e:
if _is_recursionerror(e):
raise StackError
raise
if unpacker._got_extradata():
raise ExtraData(ret, unpacker._get_extradata())
return ret
if sys.version_info < (2, 7, 6):
def _unpack_from(f, b, o=0):
return struct.unpack_from(f, bytes(b), o)
else:
_unpack_from = struct.unpack_from
_NO_FORMAT_USED = ""
_MSGPACK_HEADERS = {
0xC4: (1, _NO_FORMAT_USED, TYPE_BIN),
0xC5: (2, ">H", TYPE_BIN),
0xC6: (4, ">I", TYPE_BIN),
0xC7: (2, "Bb", TYPE_EXT),
0xC8: (3, ">Hb", TYPE_EXT),
0xC9: (5, ">Ib", TYPE_EXT),
0xCA: (4, ">f"),
0xCB: (8, ">d"),
0xCC: (1, _NO_FORMAT_USED),
0xCD: (2, ">H"),
0xCE: (4, ">I"),
0xCF: (8, ">Q"),
0xD0: (1, "b"),
0xD1: (2, ">h"),
0xD2: (4, ">i"),
0xD3: (8, ">q"),
0xD4: (1, "b1s", TYPE_EXT),
0xD5: (2, "b2s", TYPE_EXT),
0xD6: (4, "b4s", TYPE_EXT),
0xD7: (8, "b8s", TYPE_EXT),
0xD8: (16, "b16s", TYPE_EXT),
0xD9: (1, _NO_FORMAT_USED, TYPE_RAW),
0xDA: (2, ">H", TYPE_RAW),
0xDB: (4, ">I", TYPE_RAW),
0xDC: (2, ">H", TYPE_ARRAY),
0xDD: (4, ">I", TYPE_ARRAY),
0xDE: (2, ">H", TYPE_MAP),
0xDF: (4, ">I", TYPE_MAP),
}
class Unpacker(object):
def __init__(
self,
file_like=None,
read_size=0,
use_list=True,
raw=False,
timestamp=0,
strict_map_key=True,
object_hook=None,
object_pairs_hook=None,
list_hook=None,
unicode_errors=None,
max_buffer_size=100 * 1024 * 1024,
ext_hook=ExtType,
max_str_len=-1,
max_bin_len=-1,
max_array_len=-1,
max_map_len=-1,
max_ext_len=-1,
):
if unicode_errors is None:
unicode_errors = "strict"
if file_like is None:
self._feeding = True
else:
if not callable(file_like.read):
raise TypeError("`file_like.read` must be callable")
self.file_like = file_like
self._feeding = False
self._buffer = bytearray()
self._buff_i = 0
self._buf_checkpoint = 0
if not max_buffer_size:
max_buffer_size = 2**31 - 1
if max_str_len == -1:
max_str_len = max_buffer_size
if max_bin_len == -1:
max_bin_len = max_buffer_size
if max_array_len == -1:
max_array_len = max_buffer_size
if max_map_len == -1:
max_map_len = max_buffer_size // 2
if max_ext_len == -1:
max_ext_len = max_buffer_size
self._max_buffer_size = max_buffer_size
if read_size > self._max_buffer_size:
raise ValueError("read_size must be smaller than max_buffer_size")
self._read_size = read_size or min(self._max_buffer_size, 16 * 1024)
self._raw = bool(raw)
self._strict_map_key = bool(strict_map_key)
self._unicode_errors = unicode_errors
self._use_list = use_list
if not (0 <= timestamp <= 3):
raise ValueError("timestamp must be 0..3")
self._timestamp = timestamp
self._list_hook = list_hook
self._object_hook = object_hook
self._object_pairs_hook = object_pairs_hook
self._ext_hook = ext_hook
self._max_str_len = max_str_len
self._max_bin_len = max_bin_len
self._max_array_len = max_array_len
self._max_map_len = max_map_len
self._max_ext_len = max_ext_len
self._stream_offset = 0
if list_hook is not None and not callable(list_hook):
raise TypeError("`list_hook` is not callable")
if object_hook is not None and not callable(object_hook):
raise TypeError("`object_hook` is not callable")
if object_pairs_hook is not None and not callable(object_pairs_hook):
raise TypeError("`object_pairs_hook` is not callable")
if object_hook is not None and object_pairs_hook is not None:
raise TypeError(
"object_pairs_hook and object_hook are mutually " "exclusive"
)
if not callable(ext_hook):
raise TypeError("`ext_hook` is not callable")
def feed(self, next_bytes):
assert self._feeding
view = _get_data_from_buffer(next_bytes)
if len(self._buffer) - self._buff_i + len(view) > self._max_buffer_size:
raise BufferFull
if self._buf_checkpoint > 0:
del self._buffer[: self._buf_checkpoint]
self._buff_i -= self._buf_checkpoint
self._buf_checkpoint = 0
self._buffer.extend(view)
def _consume(self):
self._stream_offset += self._buff_i - self._buf_checkpoint
self._buf_checkpoint = self._buff_i
def _got_extradata(self):
return self._buff_i < len(self._buffer)
def _get_extradata(self):
return self._buffer[self._buff_i :]
def read_bytes(self, n):
ret = self._read(n, raise_outofdata=False)
self._consume()
return ret
def _read(self, n, raise_outofdata=True):
self._reserve(n, raise_outofdata=raise_outofdata)
i = self._buff_i
ret = self._buffer[i : i + n]
self._buff_i = i + len(ret)
return ret
def _reserve(self, n, raise_outofdata=True):
remain_bytes = len(self._buffer) - self._buff_i - n
if remain_bytes >= 0:
return
if self._feeding:
self._buff_i = self._buf_checkpoint
raise OutOfData
if self._buf_checkpoint > 0:
del self._buffer[: self._buf_checkpoint]
self._buff_i -= self._buf_checkpoint
self._buf_checkpoint = 0
remain_bytes = -remain_bytes
if remain_bytes + len(self._buffer) > self._max_buffer_size:
raise BufferFull
while remain_bytes > 0:
to_read_bytes = max(self._read_size, remain_bytes)
read_data = self.file_like.read(to_read_bytes)
if not read_data:
break
assert isinstance(read_data, bytes)
self._buffer += read_data
remain_bytes -= len(read_data)
if len(self._buffer) < n + self._buff_i and raise_outofdata:
self._buff_i = 0 raise OutOfData
def _read_header(self):
typ = TYPE_IMMEDIATE
n = 0
obj = None
self._reserve(1)
b = self._buffer[self._buff_i]
self._buff_i += 1
if b & 0b10000000 == 0:
obj = b
elif b & 0b11100000 == 0b11100000:
obj = -1 - (b ^ 0xFF)
elif b & 0b11100000 == 0b10100000:
n = b & 0b00011111
typ = TYPE_RAW
if n > self._max_str_len:
raise ValueError("%s exceeds max_str_len(%s)" % (n, self._max_str_len))
obj = self._read(n)
elif b & 0b11110000 == 0b10010000:
n = b & 0b00001111
typ = TYPE_ARRAY
if n > self._max_array_len:
raise ValueError(
"%s exceeds max_array_len(%s)" % (n, self._max_array_len)
)
elif b & 0b11110000 == 0b10000000:
n = b & 0b00001111
typ = TYPE_MAP
if n > self._max_map_len:
raise ValueError("%s exceeds max_map_len(%s)" % (n, self._max_map_len))
elif b == 0xC0:
obj = None
elif b == 0xC2:
obj = False
elif b == 0xC3:
obj = True
elif 0xC4 <= b <= 0xC6:
size, fmt, typ = _MSGPACK_HEADERS[b]
self._reserve(size)
if len(fmt) > 0:
n = _unpack_from(fmt, self._buffer, self._buff_i)[0]
else:
n = self._buffer[self._buff_i]
self._buff_i += size
if n > self._max_bin_len:
raise ValueError("%s exceeds max_bin_len(%s)" % (n, self._max_bin_len))
obj = self._read(n)
elif 0xC7 <= b <= 0xC9:
size, fmt, typ = _MSGPACK_HEADERS[b]
self._reserve(size)
L, n = _unpack_from(fmt, self._buffer, self._buff_i)
self._buff_i += size
if L > self._max_ext_len:
raise ValueError("%s exceeds max_ext_len(%s)" % (L, self._max_ext_len))
obj = self._read(L)
elif 0xCA <= b <= 0xD3:
size, fmt = _MSGPACK_HEADERS[b]
self._reserve(size)
if len(fmt) > 0:
obj = _unpack_from(fmt, self._buffer, self._buff_i)[0]
else:
obj = self._buffer[self._buff_i]
self._buff_i += size
elif 0xD4 <= b <= 0xD8:
size, fmt, typ = _MSGPACK_HEADERS[b]
if self._max_ext_len < size:
raise ValueError(
"%s exceeds max_ext_len(%s)" % (size, self._max_ext_len)
)
self._reserve(size + 1)
n, obj = _unpack_from(fmt, self._buffer, self._buff_i)
self._buff_i += size + 1
elif 0xD9 <= b <= 0xDB:
size, fmt, typ = _MSGPACK_HEADERS[b]
self._reserve(size)
if len(fmt) > 0:
(n,) = _unpack_from(fmt, self._buffer, self._buff_i)
else:
n = self._buffer[self._buff_i]
self._buff_i += size
if n > self._max_str_len:
raise ValueError("%s exceeds max_str_len(%s)" % (n, self._max_str_len))
obj = self._read(n)
elif 0xDC <= b <= 0xDD:
size, fmt, typ = _MSGPACK_HEADERS[b]
self._reserve(size)
(n,) = _unpack_from(fmt, self._buffer, self._buff_i)
self._buff_i += size
if n > self._max_array_len:
raise ValueError(
"%s exceeds max_array_len(%s)" % (n, self._max_array_len)
)
elif 0xDE <= b <= 0xDF:
size, fmt, typ = _MSGPACK_HEADERS[b]
self._reserve(size)
(n,) = _unpack_from(fmt, self._buffer, self._buff_i)
self._buff_i += size
if n > self._max_map_len:
raise ValueError("%s exceeds max_map_len(%s)" % (n, self._max_map_len))
else:
raise FormatError("Unknown header: 0x%x" % b)
return typ, n, obj
def _unpack(self, execute=EX_CONSTRUCT):
typ, n, obj = self._read_header()
if execute == EX_READ_ARRAY_HEADER:
if typ != TYPE_ARRAY:
raise ValueError("Expected array")
return n
if execute == EX_READ_MAP_HEADER:
if typ != TYPE_MAP:
raise ValueError("Expected map")
return n
if typ == TYPE_ARRAY:
if execute == EX_SKIP:
for i in xrange(n):
self._unpack(EX_SKIP)
return
ret = newlist_hint(n)
for i in xrange(n):
ret.append(self._unpack(EX_CONSTRUCT))
if self._list_hook is not None:
ret = self._list_hook(ret)
return ret if self._use_list else tuple(ret)
if typ == TYPE_MAP:
if execute == EX_SKIP:
for i in xrange(n):
self._unpack(EX_SKIP)
self._unpack(EX_SKIP)
return
if self._object_pairs_hook is not None:
ret = self._object_pairs_hook(
(self._unpack(EX_CONSTRUCT), self._unpack(EX_CONSTRUCT))
for _ in xrange(n)
)
else:
ret = {}
for _ in xrange(n):
key = self._unpack(EX_CONSTRUCT)
if self._strict_map_key and type(key) not in (unicode, bytes):
raise ValueError(
"%s is not allowed for map key" % str(type(key))
)
if not PY2 and type(key) is str:
key = sys.intern(key)
ret[key] = self._unpack(EX_CONSTRUCT)
if self._object_hook is not None:
ret = self._object_hook(ret)
return ret
if execute == EX_SKIP:
return
if typ == TYPE_RAW:
if self._raw:
obj = bytes(obj)
else:
obj = obj.decode("utf_8", self._unicode_errors)
return obj
if typ == TYPE_BIN:
return bytes(obj)
if typ == TYPE_EXT:
if n == -1: ts = Timestamp.from_bytes(bytes(obj))
if self._timestamp == 1:
return ts.to_unix()
elif self._timestamp == 2:
return ts.to_unix_nano()
elif self._timestamp == 3:
return ts.to_datetime()
else:
return ts
else:
return self._ext_hook(n, bytes(obj))
assert typ == TYPE_IMMEDIATE
return obj
def __iter__(self):
return self
def __next__(self):
try:
ret = self._unpack(EX_CONSTRUCT)
self._consume()
return ret
except OutOfData:
self._consume()
raise StopIteration
except RecursionError:
raise StackError
next = __next__
def skip(self):
self._unpack(EX_SKIP)
self._consume()
def unpack(self):
try:
ret = self._unpack(EX_CONSTRUCT)
except RecursionError:
raise StackError
self._consume()
return ret
def read_array_header(self):
ret = self._unpack(EX_READ_ARRAY_HEADER)
self._consume()
return ret
def read_map_header(self):
ret = self._unpack(EX_READ_MAP_HEADER)
self._consume()
return ret
def tell(self):
return self._stream_offset
class Packer(object):
def __init__(
self,
default=None,
use_single_float=False,
autoreset=True,
use_bin_type=True,
strict_types=False,
datetime=False,
unicode_errors=None,
):
self._strict_types = strict_types
self._use_float = use_single_float
self._autoreset = autoreset
self._use_bin_type = use_bin_type
self._buffer = StringIO()
if PY2 and datetime:
raise ValueError("datetime is not supported in Python 2")
self._datetime = bool(datetime)
self._unicode_errors = unicode_errors or "strict"
if default is not None:
if not callable(default):
raise TypeError("default must be callable")
self._default = default
def _pack(
self,
obj,
nest_limit=DEFAULT_RECURSE_LIMIT,
check=isinstance,
check_type_strict=_check_type_strict,
):
default_used = False
if self._strict_types:
check = check_type_strict
list_types = list
else:
list_types = (list, tuple)
while True:
if nest_limit < 0:
raise ValueError("recursion limit exceeded")
if obj is None:
return self._buffer.write(b"\xc0")
if check(obj, bool):
if obj:
return self._buffer.write(b"\xc3")
return self._buffer.write(b"\xc2")
if check(obj, int_types):
if 0 <= obj < 0x80:
return self._buffer.write(struct.pack("B", obj))
if -0x20 <= obj < 0:
return self._buffer.write(struct.pack("b", obj))
if 0x80 <= obj <= 0xFF:
return self._buffer.write(struct.pack("BB", 0xCC, obj))
if -0x80 <= obj < 0:
return self._buffer.write(struct.pack(">Bb", 0xD0, obj))
if 0xFF < obj <= 0xFFFF:
return self._buffer.write(struct.pack(">BH", 0xCD, obj))
if -0x8000 <= obj < -0x80:
return self._buffer.write(struct.pack(">Bh", 0xD1, obj))
if 0xFFFF < obj <= 0xFFFFFFFF:
return self._buffer.write(struct.pack(">BI", 0xCE, obj))
if -0x80000000 <= obj < -0x8000:
return self._buffer.write(struct.pack(">Bi", 0xD2, obj))
if 0xFFFFFFFF < obj <= 0xFFFFFFFFFFFFFFFF:
return self._buffer.write(struct.pack(">BQ", 0xCF, obj))
if -0x8000000000000000 <= obj < -0x80000000:
return self._buffer.write(struct.pack(">Bq", 0xD3, obj))
if not default_used and self._default is not None:
obj = self._default(obj)
default_used = True
continue
raise OverflowError("Integer value out of range")
if check(obj, (bytes, bytearray)):
n = len(obj)
if n >= 2**32:
raise ValueError("%s is too large" % type(obj).__name__)
self._pack_bin_header(n)
return self._buffer.write(obj)
if check(obj, unicode):
obj = obj.encode("utf-8", self._unicode_errors)
n = len(obj)
if n >= 2**32:
raise ValueError("String is too large")
self._pack_raw_header(n)
return self._buffer.write(obj)
if check(obj, memoryview):
n = obj.nbytes
if n >= 2**32:
raise ValueError("Memoryview is too large")
self._pack_bin_header(n)
return self._buffer.write(obj)
if check(obj, float):
if self._use_float:
return self._buffer.write(struct.pack(">Bf", 0xCA, obj))
return self._buffer.write(struct.pack(">Bd", 0xCB, obj))
if check(obj, (ExtType, Timestamp)):
if check(obj, Timestamp):
code = -1
data = obj.to_bytes()
else:
code = obj.code
data = obj.data
assert isinstance(code, int)
assert isinstance(data, bytes)
L = len(data)
if L == 1:
self._buffer.write(b"\xd4")
elif L == 2:
self._buffer.write(b"\xd5")
elif L == 4:
self._buffer.write(b"\xd6")
elif L == 8:
self._buffer.write(b"\xd7")
elif L == 16:
self._buffer.write(b"\xd8")
elif L <= 0xFF:
self._buffer.write(struct.pack(">BB", 0xC7, L))
elif L <= 0xFFFF:
self._buffer.write(struct.pack(">BH", 0xC8, L))
else:
self._buffer.write(struct.pack(">BI", 0xC9, L))
self._buffer.write(struct.pack("b", code))
self._buffer.write(data)
return
if check(obj, list_types):
n = len(obj)
self._pack_array_header(n)
for i in xrange(n):
self._pack(obj[i], nest_limit - 1)
return
if check(obj, dict):
return self._pack_map_pairs(
len(obj), dict_iteritems(obj), nest_limit - 1
)
if self._datetime and check(obj, _DateTime) and obj.tzinfo is not None:
obj = Timestamp.from_datetime(obj)
default_used = 1
continue
if not default_used and self._default is not None:
obj = self._default(obj)
default_used = 1
continue
if self._datetime and check(obj, _DateTime):
raise ValueError("Cannot serialize %r where tzinfo=None" % (obj,))
raise TypeError("Cannot serialize %r" % (obj,))
def pack(self, obj):
try:
self._pack(obj)
except:
self._buffer = StringIO() raise
if self._autoreset:
ret = self._buffer.getvalue()
self._buffer = StringIO()
return ret
def pack_map_pairs(self, pairs):
self._pack_map_pairs(len(pairs), pairs)
if self._autoreset:
ret = self._buffer.getvalue()
self._buffer = StringIO()
return ret
def pack_array_header(self, n):
if n >= 2**32:
raise ValueError
self._pack_array_header(n)
if self._autoreset:
ret = self._buffer.getvalue()
self._buffer = StringIO()
return ret
def pack_map_header(self, n):
if n >= 2**32:
raise ValueError
self._pack_map_header(n)
if self._autoreset:
ret = self._buffer.getvalue()
self._buffer = StringIO()
return ret
def pack_ext_type(self, typecode, data):
if not isinstance(typecode, int):
raise TypeError("typecode must have int type.")
if not 0 <= typecode <= 127:
raise ValueError("typecode should be 0-127")
if not isinstance(data, bytes):
raise TypeError("data must have bytes type")
L = len(data)
if L > 0xFFFFFFFF:
raise ValueError("Too large data")
if L == 1:
self._buffer.write(b"\xd4")
elif L == 2:
self._buffer.write(b"\xd5")
elif L == 4:
self._buffer.write(b"\xd6")
elif L == 8:
self._buffer.write(b"\xd7")
elif L == 16:
self._buffer.write(b"\xd8")
elif L <= 0xFF:
self._buffer.write(b"\xc7" + struct.pack("B", L))
elif L <= 0xFFFF:
self._buffer.write(b"\xc8" + struct.pack(">H", L))
else:
self._buffer.write(b"\xc9" + struct.pack(">I", L))
self._buffer.write(struct.pack("B", typecode))
self._buffer.write(data)
def _pack_array_header(self, n):
if n <= 0x0F:
return self._buffer.write(struct.pack("B", 0x90 + n))
if n <= 0xFFFF:
return self._buffer.write(struct.pack(">BH", 0xDC, n))
if n <= 0xFFFFFFFF:
return self._buffer.write(struct.pack(">BI", 0xDD, n))
raise ValueError("Array is too large")
def _pack_map_header(self, n):
if n <= 0x0F:
return self._buffer.write(struct.pack("B", 0x80 + n))
if n <= 0xFFFF:
return self._buffer.write(struct.pack(">BH", 0xDE, n))
if n <= 0xFFFFFFFF:
return self._buffer.write(struct.pack(">BI", 0xDF, n))
raise ValueError("Dict is too large")
def _pack_map_pairs(self, n, pairs, nest_limit=DEFAULT_RECURSE_LIMIT):
self._pack_map_header(n)
for (k, v) in pairs:
self._pack(k, nest_limit - 1)
self._pack(v, nest_limit - 1)
def _pack_raw_header(self, n):
if n <= 0x1F:
self._buffer.write(struct.pack("B", 0xA0 + n))
elif self._use_bin_type and n <= 0xFF:
self._buffer.write(struct.pack(">BB", 0xD9, n))
elif n <= 0xFFFF:
self._buffer.write(struct.pack(">BH", 0xDA, n))
elif n <= 0xFFFFFFFF:
self._buffer.write(struct.pack(">BI", 0xDB, n))
else:
raise ValueError("Raw is too large")
def _pack_bin_header(self, n):
if not self._use_bin_type:
return self._pack_raw_header(n)
elif n <= 0xFF:
return self._buffer.write(struct.pack(">BB", 0xC4, n))
elif n <= 0xFFFF:
return self._buffer.write(struct.pack(">BH", 0xC5, n))
elif n <= 0xFFFFFFFF:
return self._buffer.write(struct.pack(">BI", 0xC6, n))
else:
raise ValueError("Bin is too large")
def bytes(self):
return self._buffer.getvalue()
def reset(self):
self._buffer = StringIO()
def getbuffer(self):
if USING_STRINGBUILDER or PY2:
return memoryview(self.bytes())
else:
return self._buffer.getbuffer()