summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorINADA Naoki <inada-n@klab.com>2014-06-26 21:41:50 +0900
committerINADA Naoki <inada-n@klab.com>2014-06-26 21:41:50 +0900
commitc3cc8a96460c94cdde8eee6e83df6cb661d42018 (patch)
tree5d53902deaa4a7a439b47c21f029744ce5909f1c
parentfc546ec4cd135467aa3ee2e33061f17a6549ae98 (diff)
downloadmsgpack-python-c3cc8a96460c94cdde8eee6e83df6cb661d42018.tar.gz
Implement max_xxx_type for fallback.
-rw-r--r--msgpack/_unpacker.pyx3
-rw-r--r--msgpack/fallback.py127
2 files changed, 99 insertions, 31 deletions
diff --git a/msgpack/_unpacker.pyx b/msgpack/_unpacker.pyx
index f69be91..a9770d4 100644
--- a/msgpack/_unpacker.pyx
+++ b/msgpack/_unpacker.pyx
@@ -163,8 +163,7 @@ def unpack(object stream, object object_hook=None, object list_hook=None,
cdef class Unpacker(object):
- """
- Streaming unpacker.
+ """Streaming unpacker.
arguments:
diff --git a/msgpack/fallback.py b/msgpack/fallback.py
index 49323e6..f50adde 100644
--- a/msgpack/fallback.py
+++ b/msgpack/fallback.py
@@ -99,62 +99,84 @@ def unpackb(packed, **kwargs):
class Unpacker(object):
- """
- Streaming unpacker.
+ """Streaming unpacker.
+
+ arguments:
- `file_like` is a file-like object having a `.read(n)` method.
- When `Unpacker` is initialized with a `file_like`, `.feed()` is not
- usable.
+ :param file_like:
+ File-like object having `.read(n)` method.
+ If specified, unpacker reads serialized data from it and :meth:`feed()` is not usable.
- `read_size` is used for `file_like.read(read_size)`.
+ :param int read_size:
+ Used as `file_like.read(read_size)`. (default: `min(1024**2, max_buffer_size)`)
- If `use_list` is True (default), msgpack lists are deserialized to Python
- lists. Otherwise they are deserialized to tuples.
+ :param bool use_list:
+ If true, unpack msgpack array to Python list.
+ Otherwise, unpack to Python tuple. (default: True)
- `object_hook` is the same as in simplejson. If it is not None, it should
- be callable and Unpacker calls it with a dict argument after deserializing
- a map.
+ :param callable object_hook:
+ When specified, it should be callable.
+ Unpacker calls it with a dict argument after unpacking msgpack map.
+ (See also simplejson)
- `object_pairs_hook` is the same as in simplejson. If it is not None, it
- should be callable and Unpacker calls it with a list of key-value pairs
- after deserializing a map.
+ :param callable object_pairs_hook:
+ When specified, it should be callable.
+ Unpacker calls it with a list of key-value pairs after unpacking msgpack map.
+ (See also simplejson)
- `ext_hook` is callback for ext (User defined) type. It called with two
- arguments: (code, bytes). default: `msgpack.ExtType`
+ :param str encoding:
+ Encoding used for decoding msgpack raw.
+ If it is None (default), msgpack raw is deserialized to Python bytes.
- `encoding` is the encoding used for decoding msgpack bytes. If it is
- None (default), msgpack bytes are deserialized to Python bytes.
+ :param str unicode_errors:
+ Used for decoding msgpack raw with *encoding*.
+ (default: `'strict'`)
- `unicode_errors` is used for decoding bytes.
+ :param int max_buffer_size:
+ Limits size of data waiting unpacked. 0 means system's INT_MAX (default).
+ Raises `BufferFull` exception when it is insufficient.
+ You shoud set this parameter when unpacking data from untrasted source.
- `max_buffer_size` limits the buffer size. 0 means INT_MAX (default).
+ :param int max_str_len:
+ Limits max length of str. (default: 2**31-1)
- Raises `BufferFull` exception when it is unsufficient.
+ :param int max_bin_len:
+ Limits max length of bin. (default: 2**31-1)
- You should set this parameter when unpacking data from an untrustred source.
+ :param int max_array_len:
+ Limits max length of array. (default: 2**31-1)
- example of streaming deserialization from file-like object::
+ :param int max_map_len:
+ Limits max length of map. (default: 2**31-1)
+
+
+ example of streaming deserialize from file-like object::
unpacker = Unpacker(file_like)
for o in unpacker:
- do_something(o)
+ process(o)
- example of streaming deserialization from socket::
+ example of streaming deserialize from socket::
unpacker = Unpacker()
- while 1:
- buf = sock.recv(1024*2)
+ while True:
+ buf = sock.recv(1024**2)
if not buf:
break
unpacker.feed(buf)
for o in unpacker:
- do_something(o)
+ process(o)
"""
def __init__(self, file_like=None, read_size=0, use_list=True,
object_hook=None, object_pairs_hook=None, list_hook=None,
encoding=None, unicode_errors='strict', max_buffer_size=0,
- ext_hook=ExtType):
+ ext_hook=ExtType,
+ max_str_len=2147483647, # 2**32-1
+ max_bin_len=2147483647,
+ max_array_len=2147483647,
+ max_map_len=2147483647,
+ max_ext_len=2147483647):
if file_like is None:
self._fb_feeding = True
else:
@@ -177,6 +199,11 @@ class Unpacker(object):
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
if list_hook is not None and not callable(list_hook):
raise TypeError('`list_hook` is not callable')
@@ -280,12 +307,18 @@ class Unpacker(object):
n = b & 0b00011111
obj = self._fb_read(n, write_bytes)
typ = TYPE_RAW
+ if n > self._max_str_len:
+ raise ValueError("%s exceeds max_str_len(%s)", n, self._max_str_len)
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:
@@ -295,26 +328,38 @@ class Unpacker(object):
elif b == 0xc4:
typ = TYPE_BIN
n = struct.unpack("B", self._fb_read(1, write_bytes))[0]
+ if n > self._max_bin_len:
+ raise ValueError("%s exceeds max_bin_len(%s)" % (n, self._max_bin_len))
obj = self._fb_read(n, write_bytes)
elif b == 0xc5:
typ = TYPE_BIN
n = struct.unpack(">H", self._fb_read(2, write_bytes))[0]
+ if n > self._max_bin_len:
+ raise ValueError("%s exceeds max_bin_len(%s)" % (n, self._max_bin_len))
obj = self._fb_read(n, write_bytes)
elif b == 0xc6:
typ = TYPE_BIN
n = struct.unpack(">I", self._fb_read(4, write_bytes))[0]
+ if n > self._max_bin_len:
+ raise ValueError("%s exceeds max_bin_len(%s)" % (n, self._max_bin_len))
obj = self._fb_read(n, write_bytes)
elif b == 0xc7: # ext 8
typ = TYPE_EXT
L, n = struct.unpack('Bb', self._fb_read(2, write_bytes))
+ if L > self._max_ext_len:
+ raise ValueError("%s exceeds max_ext_len(%s)" % (L, self._max_ext_len))
obj = self._fb_read(L, write_bytes)
elif b == 0xc8: # ext 16
typ = TYPE_EXT
L, n = struct.unpack('>Hb', self._fb_read(3, write_bytes))
+ if L > self._max_ext_len:
+ raise ValueError("%s exceeds max_ext_len(%s)" % (L, self._max_ext_len))
obj = self._fb_read(L, write_bytes)
elif b == 0xc9: # ext 32
typ = TYPE_EXT
L, n = struct.unpack('>Ib', self._fb_read(5, write_bytes))
+ if L > self._max_ext_len:
+ raise ValueError("%s exceeds max_ext_len(%s)" % (L, self._max_ext_len))
obj = self._fb_read(L, write_bytes)
elif b == 0xca:
obj = struct.unpack(">f", self._fb_read(4, write_bytes))[0]
@@ -338,42 +383,66 @@ class Unpacker(object):
obj = struct.unpack(">q", self._fb_read(8, write_bytes))[0]
elif b == 0xd4: # fixext 1
typ = TYPE_EXT
+ if self._max_ext_len < 1:
+ raise ValueError("%s exceeds max_ext_len(%s)" % (1, self._max_ext_len))
n, obj = struct.unpack('b1s', self._fb_read(2, write_bytes))
elif b == 0xd5: # fixext 2
typ = TYPE_EXT
+ if self._max_ext_len < 2:
+ raise ValueError("%s exceeds max_ext_len(%s)" % (2, self._max_ext_len))
n, obj = struct.unpack('b2s', self._fb_read(3, write_bytes))
elif b == 0xd6: # fixext 4
typ = TYPE_EXT
+ if self._max_ext_len < 4:
+ raise ValueError("%s exceeds max_ext_len(%s)" % (4, self._max_ext_len))
n, obj = struct.unpack('b4s', self._fb_read(5, write_bytes))
elif b == 0xd7: # fixext 8
typ = TYPE_EXT
+ if self._max_ext_len < 8:
+ raise ValueError("%s exceeds max_ext_len(%s)" % (8, self._max_ext_len))
n, obj = struct.unpack('b8s', self._fb_read(9, write_bytes))
elif b == 0xd8: # fixext 16
typ = TYPE_EXT
+ if self._max_ext_len < 16:
+ raise ValueError("%s exceeds max_ext_len(%s)" % (16, self._max_ext_len))
n, obj = struct.unpack('b16s', self._fb_read(17, write_bytes))
elif b == 0xd9:
typ = TYPE_RAW
n = struct.unpack("B", self._fb_read(1, write_bytes))[0]
+ if n > self._max_str_len:
+ raise ValueError("%s exceeds max_str_len(%s)", n, self._max_str_len)
obj = self._fb_read(n, write_bytes)
elif b == 0xda:
typ = TYPE_RAW
n = struct.unpack(">H", self._fb_read(2, write_bytes))[0]
+ if n > self._max_str_len:
+ raise ValueError("%s exceeds max_str_len(%s)", n, self._max_str_len)
obj = self._fb_read(n, write_bytes)
elif b == 0xdb:
typ = TYPE_RAW
n = struct.unpack(">I", self._fb_read(4, write_bytes))[0]
+ if n > self._max_str_len:
+ raise ValueError("%s exceeds max_str_len(%s)", n, self._max_str_len)
obj = self._fb_read(n, write_bytes)
elif b == 0xdc:
n = struct.unpack(">H", self._fb_read(2, write_bytes))[0]
+ if n > self._max_array_len:
+ raise ValueError("%s exceeds max_array_len(%s)", n, self._max_array_len)
typ = TYPE_ARRAY
elif b == 0xdd:
n = struct.unpack(">I", self._fb_read(4, write_bytes))[0]
+ if n > self._max_array_len:
+ raise ValueError("%s exceeds max_array_len(%s)", n, self._max_array_len)
typ = TYPE_ARRAY
elif b == 0xde:
n = struct.unpack(">H", self._fb_read(2, write_bytes))[0]
+ if n > self._max_map_len:
+ raise ValueError("%s exceeds max_map_len(%s)", n, self._max_map_len)
typ = TYPE_MAP
elif b == 0xdf:
n = struct.unpack(">I", self._fb_read(4, write_bytes))[0]
+ if n > self._max_map_len:
+ raise ValueError("%s exceeds max_map_len(%s)", n, self._max_map_len)
typ = TYPE_MAP
else:
raise UnpackValueError("Unknown header: 0x%x" % b)