aboutsummaryrefslogtreecommitdiff
path: root/python/olm
diff options
context:
space:
mode:
authorDamir Jelić <poljar@termina.org.uk>2019-06-18 13:46:57 +0200
committerDamir Jelić <poljar@termina.org.uk>2019-06-18 13:50:46 +0200
commit2f5590bf38e5995a36f770c04cfbf31eb9070eca (patch)
tree299e68cd3b9ea33f212cd1a7c4fcf0bbbca4bfc8 /python/olm
parente1a4e6ebf1568935a57ba8cec48e43dd7c1ebcd3 (diff)
olm: Allow decryption functions to define how to handle unicode decode errors.
This patch changes the decryption functions not to fail if there was an unicode decode error while converting the decrypted bytes plaintext into a native python string. Characters that cannot be decoded as unicode are now replaced with the unicode replacement character (U+FFFD). The old behaviour of raising an UnicodeDecodeError can be achieved by passing the "strict" error handling scheme to the decrypt function.
Diffstat (limited to 'python/olm')
-rw-r--r--python/olm/group_session.py21
-rw-r--r--python/olm/pk.py15
-rw-r--r--python/olm/sas.py4
-rw-r--r--python/olm/session.py21
-rw-r--r--python/olm/utility.py1
5 files changed, 43 insertions, 19 deletions
diff --git a/python/olm/group_session.py b/python/olm/group_session.py
index 737d9ef..88f87f0 100644
--- a/python/olm/group_session.py
+++ b/python/olm/group_session.py
@@ -33,7 +33,7 @@ from future.utils import bytes_to_native_str
# pylint: disable=no-name-in-module
from _libolm import ffi, lib # type: ignore
-from ._compat import URANDOM, to_bytearray, to_bytes
+from ._compat import URANDOM, to_bytearray, to_bytes, to_native_str
from ._finalize import track_for_finalization
@@ -176,8 +176,8 @@ class InboundGroupSession(object):
raise OlmGroupSessionError(last_error)
- def decrypt(self, ciphertext):
- # type: (AnyStr) -> Tuple[str, int]
+ def decrypt(self, ciphertext, errors="replace"):
+ # type: (AnyStr, str) -> Tuple[str, int]
"""Decrypt a message
Returns a tuple of the decrypted plain-text and the message index of
@@ -197,6 +197,13 @@ class InboundGroupSession(object):
Args:
ciphertext(str): Base64 encoded ciphertext containing the encrypted
message
+ unicode_errors(str, optional): The error handling scheme to use for
+ unicode decoding errors. The default is "replace" meaning that
+ the character that was unable to decode will be replaced with
+ the unicode replacement character (U+FFFD). Other possible
+ values are "strict", "ignore" and "xmlcharrefreplace" as well
+ as any other name registered with codecs.register_error that
+ can handle UnicodeEncodeErrors.
"""
if not ciphertext:
raise ValueError("Ciphertext can't be empty.")
@@ -223,10 +230,10 @@ class InboundGroupSession(object):
self._check_error(plaintext_length)
- plaintext = bytes_to_native_str(ffi.unpack(
- plaintext_buffer,
- plaintext_length
- ))
+ plaintext = to_native_str(
+ ffi.unpack(plaintext_buffer, plaintext_length),
+ errors=errors
+ )
# clear out copies of the plaintext
lib.memset(plaintext_buffer, 0, max_plaintext_length)
diff --git a/python/olm/pk.py b/python/olm/pk.py
index 193aba5..158c78d 100644
--- a/python/olm/pk.py
+++ b/python/olm/pk.py
@@ -40,7 +40,7 @@ from future.utils import bytes_to_native_str
from _libolm import ffi, lib # type: ignore
-from ._compat import URANDOM, to_bytearray
+from ._compat import URANDOM, to_bytearray, to_native_str
from ._finalize import track_for_finalization
@@ -313,8 +313,8 @@ class PkDecryption(object):
return obj
- def decrypt(self, message):
- # type (PkMessage) -> str
+ def decrypt(self, message, errors="replace"):
+ # type (PkMessage, str) -> str
"""Decrypt a previously encrypted Pk message.
Returns the decrypted plaintext.
@@ -322,6 +322,13 @@ class PkDecryption(object):
Args:
message(PkMessage): the pk message to decrypt.
+ unicode_errors(str, optional): The error handling scheme to use for
+ unicode decoding errors. The default is "replace" meaning that
+ the character that was unable to decode will be replaced with
+ the unicode replacement character (U+FFFD). Other possible
+ values are "strict", "ignore" and "xmlcharrefreplace" as well
+ as any other name registered with codecs.register_error that
+ can handle UnicodeEncodeErrors.
"""
ephemeral_key = to_bytearray(message.ephemeral_key)
ephemeral_key_size = len(ephemeral_key)
@@ -354,7 +361,7 @@ class PkDecryption(object):
# clear out copies of the plaintext
lib.memset(plaintext_buffer, 0, max_plaintext_length)
- return bytes_to_native_str(plaintext)
+ return to_native_str(plaintext, errors=errors)
def _clear_pk_signing(pk_struct):
diff --git a/python/olm/sas.py b/python/olm/sas.py
index c12b7bc..bea1dd0 100644
--- a/python/olm/sas.py
+++ b/python/olm/sas.py
@@ -30,15 +30,15 @@ Examples:
"""
-from functools import wraps
from builtins import bytes
+from functools import wraps
from typing import Optional
from future.utils import bytes_to_native_str
from _libolm import ffi, lib
-from ._compat import URANDOM, to_bytes, to_bytearray
+from ._compat import URANDOM, to_bytearray, to_bytes
from ._finalize import track_for_finalization
diff --git a/python/olm/session.py b/python/olm/session.py
index cba9be0..cf66582 100644
--- a/python/olm/session.py
+++ b/python/olm/session.py
@@ -40,7 +40,7 @@ from future.utils import bytes_to_native_str
# pylint: disable=no-name-in-module
from _libolm import ffi, lib # type: ignore
-from ._compat import URANDOM, to_bytearray, to_bytes
+from ._compat import URANDOM, to_bytearray, to_bytes, to_native_str
from ._finalize import track_for_finalization
# This is imported only for type checking purposes
@@ -273,8 +273,8 @@ class Session(object):
else: # pragma: no cover
raise ValueError("Unknown message type")
- def decrypt(self, message):
- # type: (_OlmMessage) -> str
+ def decrypt(self, message, errors="replace"):
+ # type: (_OlmMessage, str) -> str
"""Decrypts a message using the session. Returns the plaintext string
on success. Raises OlmSessionError on failure. If the base64 couldn't
be decoded then the error message will be "INVALID_BASE64". If the
@@ -285,7 +285,14 @@ class Session(object):
Args:
message(OlmMessage): The Olm message that will be decrypted. It can
- be either an OlmPreKeyMessage or an OlmMessage.
+ be either an OlmPreKeyMessage or an OlmMessage.
+ unicode_errors(str, optional): The error handling scheme to use for
+ unicode decoding errors. The default is "replace" meaning that
+ the character that was unable to decode will be replaced with
+ the unicode replacement character (U+FFFD). Other possible
+ values are "strict", "ignore" and "xmlcharrefreplace" as well
+ as any other name registered with codecs.register_error that
+ can handle UnicodeEncodeErrors.
"""
if not message.ciphertext:
raise ValueError("Ciphertext can't be empty")
@@ -311,8 +318,10 @@ class Session(object):
plaintext_buffer, max_plaintext_length
)
self._check_error(plaintext_length)
- plaintext = bytes_to_native_str(
- ffi.unpack(plaintext_buffer, plaintext_length))
+ plaintext = to_native_str(
+ ffi.unpack(plaintext_buffer, plaintext_length),
+ errors=errors
+ )
# clear out copies of the plaintext
lib.memset(plaintext_buffer, 0, max_plaintext_length)
diff --git a/python/olm/utility.py b/python/olm/utility.py
index 10d5ab4..bddef38 100644
--- a/python/olm/utility.py
+++ b/python/olm/utility.py
@@ -32,6 +32,7 @@ Examples:
# pylint: disable=redefined-builtin,unused-import
from typing import AnyStr, Type
+
from future.utils import bytes_to_native_str
# pylint: disable=no-name-in-module