def _get_int(data):
try:
value = struct.unpack("<i", data[:4])[0]
except struct.error:
raise InvalidBSON()
return (value, data[4:])
def _get_c_string(data, length=None):
if length is None:
try:
length = data.index("\x00")
except ValueError:
raise InvalidBSON()
return (unicode(data[:length], "utf-8"), data[length + 1:])
def _get_binary(data):
(length, data) = _get_int(data)
subtype = ord(data[0])
data = data[1:]
if subtype == 2:
(length2, data) = _get_int(data)
if length2 != length - 4:
raise InvalidBSON("invalid binary (st 2) - lengths don't match!")
length = length2
if subtype == 3 and _use_uuid:
return (uuid.UUID(bytes=data[:length]), data[length:])
return (Binary(data[:length], subtype), data[length:])
def _validate_document(data, valid_name=None):
try:
obj_size = struct.unpack("<i", data[:4])[0]
except struct.error:
raise InvalidBSON()
assert obj_size <= len(data)
obj = data[4:obj_size]
assert len(obj)
eoo = obj[-1]
assert eoo == "\x00"
elements = obj[:-1]
_validate_elements(elements, valid_name)
return data[obj_size:]
def is_valid(bson):
"""Validate that the given string represents valid BSON data.
Raises TypeError if the data is not an instance of a subclass of str.
Returns True if the data represents a valid BSON object, False otherwise.
:Parameters:
- `bson`: the data to be validated
"""
if not isinstance(bson, types.StringType):
raise TypeError("BSON data must be an instance of a subclass of str")
# 4 MB limit
if len(bson) > 4 * 1024 * 1024:
raise InvalidBSON("BSON documents are limited to 4MB")
try:
remainder = _validate_document(bson)
return remainder == ""
except (AssertionError, InvalidBSON):
return False
def _validate_element_data(type, data):
try:
return _element_validator[type](data)
except KeyError:
raise InvalidBSON("unrecognized type: %s" % type)