2016-01-06 21:26:29 -07:00
|
|
|
# Copyright 2014-2016 OpenMarket Ltd
|
2021-11-23 04:43:56 -07:00
|
|
|
# Copyright 2021 The Matrix.org Foundation C.I.C.
|
2014-12-03 09:07:21 -07:00
|
|
|
#
|
|
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
|
|
# you may not use this file except in compliance with the License.
|
|
|
|
# You may obtain a copy of the License at
|
|
|
|
#
|
|
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
|
|
#
|
|
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
|
|
# See the License for the specific language governing permissions and
|
|
|
|
# limitations under the License.
|
2020-07-20 11:33:04 -06:00
|
|
|
import collections.abc
|
2018-07-09 00:09:20 -06:00
|
|
|
import re
|
2022-01-26 06:27:04 -07:00
|
|
|
from typing import (
|
|
|
|
TYPE_CHECKING,
|
|
|
|
Any,
|
|
|
|
Callable,
|
|
|
|
Dict,
|
|
|
|
Iterable,
|
|
|
|
List,
|
|
|
|
Mapping,
|
|
|
|
Optional,
|
|
|
|
Union,
|
|
|
|
)
|
2018-07-09 00:09:20 -06:00
|
|
|
|
2016-11-21 10:52:45 -07:00
|
|
|
from frozendict import frozendict
|
|
|
|
|
2021-08-09 19:23:31 -06:00
|
|
|
from synapse.api.constants import EventContentFields, EventTypes, RelationTypes
|
2020-05-14 11:24:01 -06:00
|
|
|
from synapse.api.errors import Codes, SynapseError
|
2020-03-09 06:58:25 -06:00
|
|
|
from synapse.api.room_versions import RoomVersion
|
2021-10-13 05:24:07 -06:00
|
|
|
from synapse.types import JsonDict
|
2021-03-17 10:51:55 -06:00
|
|
|
from synapse.util.frozenutils import unfreeze
|
2016-11-21 10:42:16 -07:00
|
|
|
|
2018-07-09 00:09:20 -06:00
|
|
|
from . import EventBase
|
2018-04-15 13:43:35 -06:00
|
|
|
|
2022-01-26 06:27:04 -07:00
|
|
|
if TYPE_CHECKING:
|
|
|
|
from synapse.storage.databases.main.relations import BundledAggregations
|
|
|
|
|
|
|
|
|
2016-11-21 10:42:16 -07:00
|
|
|
# Split strings on "." but not "\." This uses a negative lookbehind assertion for '\'
|
|
|
|
# (?<!stuff) matches if the current position in the string is not preceded
|
|
|
|
# by a match for 'stuff'.
|
|
|
|
# TODO: This is fast, but fails to handle "foo\\.bar" which should be treated as
|
|
|
|
# the literal fields "foo\" and "bar" but will instead be treated as "foo\\.bar"
|
|
|
|
SPLIT_FIELD_REGEX = re.compile(r"(?<!\\)\.")
|
|
|
|
|
2021-08-26 10:07:58 -06:00
|
|
|
CANONICALJSON_MAX_INT = (2 ** 53) - 1
|
|
|
|
CANONICALJSON_MIN_INT = -CANONICALJSON_MAX_INT
|
|
|
|
|
2014-12-03 09:07:21 -07:00
|
|
|
|
2020-03-05 08:46:44 -07:00
|
|
|
def prune_event(event: EventBase) -> EventBase:
|
2014-12-03 09:07:21 -07:00
|
|
|
"""Returns a pruned version of the given event, which removes all keys we
|
|
|
|
don't know about or think could potentially be dodgy.
|
|
|
|
|
|
|
|
This is used when we "redact" an event. We want to remove all fields that
|
|
|
|
the user has specified, but we do want to keep necessary information like
|
|
|
|
type, state_key etc.
|
2019-01-28 09:42:10 -07:00
|
|
|
"""
|
2020-03-09 06:58:25 -06:00
|
|
|
pruned_event_dict = prune_event_dict(event.room_version, event.get_dict())
|
2019-01-28 09:42:10 -07:00
|
|
|
|
2020-03-05 08:46:44 -07:00
|
|
|
from . import make_event_from_dict
|
2019-06-20 03:32:02 -06:00
|
|
|
|
2020-03-05 08:46:44 -07:00
|
|
|
pruned_event = make_event_from_dict(
|
|
|
|
pruned_event_dict, event.room_version, event.internal_metadata.get_dict()
|
2019-01-28 09:42:10 -07:00
|
|
|
)
|
|
|
|
|
2020-10-05 07:43:14 -06:00
|
|
|
# copy the internal fields
|
|
|
|
pruned_event.internal_metadata.stream_ordering = (
|
|
|
|
event.internal_metadata.stream_ordering
|
|
|
|
)
|
|
|
|
|
2021-03-17 06:33:18 -06:00
|
|
|
pruned_event.internal_metadata.outlier = event.internal_metadata.outlier
|
|
|
|
|
2019-07-18 07:41:42 -06:00
|
|
|
# Mark the event as redacted
|
|
|
|
pruned_event.internal_metadata.redacted = True
|
|
|
|
|
|
|
|
return pruned_event
|
|
|
|
|
2019-01-28 09:42:10 -07:00
|
|
|
|
2021-10-13 05:24:07 -06:00
|
|
|
def prune_event_dict(room_version: RoomVersion, event_dict: JsonDict) -> JsonDict:
|
2019-01-28 09:42:10 -07:00
|
|
|
"""Redacts the event_dict in the same way as `prune_event`, except it
|
|
|
|
operates on dicts rather than event objects
|
|
|
|
|
|
|
|
Returns:
|
2020-03-09 06:58:25 -06:00
|
|
|
A copy of the pruned event dict
|
2014-12-03 09:07:21 -07:00
|
|
|
"""
|
|
|
|
|
|
|
|
allowed_keys = [
|
|
|
|
"event_id",
|
|
|
|
"sender",
|
|
|
|
"room_id",
|
|
|
|
"hashes",
|
|
|
|
"signatures",
|
|
|
|
"content",
|
|
|
|
"type",
|
|
|
|
"state_key",
|
|
|
|
"depth",
|
|
|
|
"prev_events",
|
|
|
|
"auth_events",
|
|
|
|
"origin",
|
|
|
|
"origin_server_ts",
|
|
|
|
]
|
|
|
|
|
2021-01-05 05:41:48 -07:00
|
|
|
# Room versions from before MSC2176 had additional allowed keys.
|
|
|
|
if not room_version.msc2176_redaction_rules:
|
|
|
|
allowed_keys.extend(["prev_state", "membership"])
|
|
|
|
|
2019-01-28 09:42:10 -07:00
|
|
|
event_type = event_dict["type"]
|
2015-01-29 09:50:23 -07:00
|
|
|
|
2014-12-03 09:07:21 -07:00
|
|
|
new_content = {}
|
|
|
|
|
2021-10-13 05:24:07 -06:00
|
|
|
def add_fields(*fields: str) -> None:
|
2014-12-03 09:07:21 -07:00
|
|
|
for field in fields:
|
2019-01-28 09:42:10 -07:00
|
|
|
if field in event_dict["content"]:
|
2015-01-29 09:50:23 -07:00
|
|
|
new_content[field] = event_dict["content"][field]
|
2014-12-03 09:07:21 -07:00
|
|
|
|
|
|
|
if event_type == EventTypes.Member:
|
|
|
|
add_fields("membership")
|
2021-09-03 10:51:15 -06:00
|
|
|
if room_version.msc3375_redaction_rules:
|
2021-09-30 09:13:59 -06:00
|
|
|
add_fields(EventContentFields.AUTHORISING_USER)
|
2014-12-03 09:07:21 -07:00
|
|
|
elif event_type == EventTypes.Create:
|
2021-01-05 05:41:48 -07:00
|
|
|
# MSC2176 rules state that create events cannot be redacted.
|
|
|
|
if room_version.msc2176_redaction_rules:
|
|
|
|
return event_dict
|
|
|
|
|
2014-12-03 09:07:21 -07:00
|
|
|
add_fields("creator")
|
|
|
|
elif event_type == EventTypes.JoinRules:
|
|
|
|
add_fields("join_rule")
|
2021-07-28 05:03:01 -06:00
|
|
|
if room_version.msc3083_join_rules:
|
|
|
|
add_fields("allow")
|
2014-12-03 09:07:21 -07:00
|
|
|
elif event_type == EventTypes.PowerLevels:
|
|
|
|
add_fields(
|
|
|
|
"users",
|
|
|
|
"users_default",
|
|
|
|
"events",
|
|
|
|
"events_default",
|
|
|
|
"state_default",
|
|
|
|
"ban",
|
|
|
|
"kick",
|
|
|
|
"redact",
|
|
|
|
)
|
2021-01-05 05:41:48 -07:00
|
|
|
|
|
|
|
if room_version.msc2176_redaction_rules:
|
|
|
|
add_fields("invite")
|
|
|
|
|
2021-07-28 09:46:37 -06:00
|
|
|
if room_version.msc2716_historical:
|
|
|
|
add_fields("historical")
|
|
|
|
|
2020-03-09 06:58:25 -06:00
|
|
|
elif event_type == EventTypes.Aliases and room_version.special_case_aliases_auth:
|
2014-12-03 09:07:21 -07:00
|
|
|
add_fields("aliases")
|
2015-07-03 03:31:17 -06:00
|
|
|
elif event_type == EventTypes.RoomHistoryVisibility:
|
2015-07-06 06:05:52 -06:00
|
|
|
add_fields("history_visibility")
|
2021-01-05 05:41:48 -07:00
|
|
|
elif event_type == EventTypes.Redaction and room_version.msc2176_redaction_rules:
|
|
|
|
add_fields("redacts")
|
2021-08-09 19:23:31 -06:00
|
|
|
elif room_version.msc2716_redactions and event_type == EventTypes.MSC2716_INSERTION:
|
2021-09-21 14:06:28 -06:00
|
|
|
add_fields(EventContentFields.MSC2716_NEXT_BATCH_ID)
|
|
|
|
elif room_version.msc2716_redactions and event_type == EventTypes.MSC2716_BATCH:
|
|
|
|
add_fields(EventContentFields.MSC2716_BATCH_ID)
|
2021-08-09 19:23:31 -06:00
|
|
|
elif room_version.msc2716_redactions and event_type == EventTypes.MSC2716_MARKER:
|
|
|
|
add_fields(EventContentFields.MSC2716_MARKER_INSERTION)
|
2014-12-03 09:07:21 -07:00
|
|
|
|
2015-01-29 09:50:23 -07:00
|
|
|
allowed_fields = {k: v for k, v in event_dict.items() if k in allowed_keys}
|
2014-12-03 09:07:21 -07:00
|
|
|
|
|
|
|
allowed_fields["content"] = new_content
|
|
|
|
|
2021-10-13 05:24:07 -06:00
|
|
|
unsigned: JsonDict = {}
|
2019-01-28 09:42:10 -07:00
|
|
|
allowed_fields["unsigned"] = unsigned
|
2014-12-11 06:25:19 -07:00
|
|
|
|
2019-01-28 09:42:10 -07:00
|
|
|
event_unsigned = event_dict.get("unsigned", {})
|
2014-12-11 06:25:19 -07:00
|
|
|
|
2019-01-28 09:42:10 -07:00
|
|
|
if "age_ts" in event_unsigned:
|
|
|
|
unsigned["age_ts"] = event_unsigned["age_ts"]
|
|
|
|
if "replaces_state" in event_unsigned:
|
|
|
|
unsigned["replaces_state"] = event_unsigned["replaces_state"]
|
|
|
|
|
|
|
|
return allowed_fields
|
2014-12-05 09:20:48 -07:00
|
|
|
|
|
|
|
|
2021-10-13 05:24:07 -06:00
|
|
|
def _copy_field(src: JsonDict, dst: JsonDict, field: List[str]) -> None:
|
2016-11-21 10:42:16 -07:00
|
|
|
"""Copy the field in 'src' to 'dst'.
|
|
|
|
|
|
|
|
For example, if src={"foo":{"bar":5}} and dst={}, and field=["foo","bar"]
|
|
|
|
then dst={"foo":{"bar":5}}.
|
|
|
|
|
|
|
|
Args:
|
2021-10-13 05:24:07 -06:00
|
|
|
src: The dict to read from.
|
|
|
|
dst: The dict to modify.
|
|
|
|
field: List of keys to drill down to in 'src'.
|
2016-11-21 10:42:16 -07:00
|
|
|
"""
|
|
|
|
if len(field) == 0: # this should be impossible
|
|
|
|
return
|
|
|
|
if len(field) == 1: # common case e.g. 'origin_server_ts'
|
|
|
|
if field[0] in src:
|
|
|
|
dst[field[0]] = src[field[0]]
|
|
|
|
return
|
|
|
|
|
|
|
|
# Else is a nested field e.g. 'content.body'
|
|
|
|
# Pop the last field as that's the key to move across and we need the
|
|
|
|
# parent dict in order to access the data. Drill down to the right dict.
|
|
|
|
key_to_move = field.pop(-1)
|
|
|
|
sub_dict = src
|
|
|
|
for sub_field in field: # e.g. sub_field => "content"
|
2016-11-22 03:39:41 -07:00
|
|
|
if sub_field in sub_dict and type(sub_dict[sub_field]) in [dict, frozendict]:
|
2016-11-21 10:42:16 -07:00
|
|
|
sub_dict = sub_dict[sub_field]
|
|
|
|
else:
|
|
|
|
return
|
|
|
|
|
|
|
|
if key_to_move not in sub_dict:
|
|
|
|
return
|
|
|
|
|
|
|
|
# Insert the key into the output dictionary, creating nested objects
|
|
|
|
# as required. We couldn't do this any earlier or else we'd need to delete
|
|
|
|
# the empty objects if the key didn't exist.
|
|
|
|
sub_out_dict = dst
|
|
|
|
for sub_field in field:
|
2016-11-22 06:42:11 -07:00
|
|
|
sub_out_dict = sub_out_dict.setdefault(sub_field, {})
|
2016-11-21 10:42:16 -07:00
|
|
|
sub_out_dict[key_to_move] = sub_dict[key_to_move]
|
|
|
|
|
|
|
|
|
2021-10-13 05:24:07 -06:00
|
|
|
def only_fields(dictionary: JsonDict, fields: List[str]) -> JsonDict:
|
2016-11-21 10:42:16 -07:00
|
|
|
"""Return a new dict with only the fields in 'dictionary' which are present
|
|
|
|
in 'fields'.
|
|
|
|
|
|
|
|
If there are no event fields specified then all fields are included.
|
2020-10-23 10:38:40 -06:00
|
|
|
The entries may include '.' characters to indicate sub-fields.
|
2016-11-21 10:42:16 -07:00
|
|
|
So ['content.body'] will include the 'body' field of the 'content' object.
|
|
|
|
A literal '.' character in a field name may be escaped using a '\'.
|
|
|
|
|
|
|
|
Args:
|
2021-10-13 05:24:07 -06:00
|
|
|
dictionary: The dictionary to read from.
|
|
|
|
fields: A list of fields to copy over. Only shallow refs are
|
2016-11-21 10:42:16 -07:00
|
|
|
taken.
|
|
|
|
Returns:
|
2021-10-13 05:24:07 -06:00
|
|
|
A new dictionary with only the given fields. If fields was empty,
|
2016-11-21 10:42:16 -07:00
|
|
|
the same dictionary is returned.
|
|
|
|
"""
|
|
|
|
if len(fields) == 0:
|
|
|
|
return dictionary
|
|
|
|
|
|
|
|
# for each field, convert it:
|
|
|
|
# ["content.body.thing\.with\.dots"] => [["content", "body", "thing\.with\.dots"]]
|
|
|
|
split_fields = [SPLIT_FIELD_REGEX.split(f) for f in fields]
|
|
|
|
|
|
|
|
# for each element of the output array of arrays:
|
2016-11-22 06:42:11 -07:00
|
|
|
# remove escaping so we can use the right key names.
|
|
|
|
split_fields[:] = [
|
|
|
|
[f.replace(r"\.", r".") for f in field_array] for field_array in split_fields
|
|
|
|
]
|
2016-11-21 10:42:16 -07:00
|
|
|
|
2021-10-13 05:24:07 -06:00
|
|
|
output: JsonDict = {}
|
2016-11-21 10:42:16 -07:00
|
|
|
for field_array in split_fields:
|
|
|
|
_copy_field(dictionary, output, field_array)
|
|
|
|
return output
|
|
|
|
|
|
|
|
|
2021-10-13 05:24:07 -06:00
|
|
|
def format_event_raw(d: JsonDict) -> JsonDict:
|
2015-01-28 19:34:35 -07:00
|
|
|
return d
|
|
|
|
|
|
|
|
|
2021-10-13 05:24:07 -06:00
|
|
|
def format_event_for_client_v1(d: JsonDict) -> JsonDict:
|
2015-11-30 10:46:35 -07:00
|
|
|
d = format_event_for_client_v2(d)
|
|
|
|
|
2015-12-01 04:14:48 -07:00
|
|
|
sender = d.get("sender")
|
|
|
|
if sender is not None:
|
|
|
|
d["user_id"] = sender
|
2015-01-28 19:34:35 -07:00
|
|
|
|
2015-11-30 10:46:35 -07:00
|
|
|
copy_keys = (
|
2015-09-10 07:25:54 -06:00
|
|
|
"age",
|
|
|
|
"redacted_because",
|
|
|
|
"replaces_state",
|
|
|
|
"prev_content",
|
|
|
|
"invite_room_state",
|
2021-06-09 12:39:51 -06:00
|
|
|
"knock_room_state",
|
2015-09-10 07:25:54 -06:00
|
|
|
)
|
2015-11-30 10:46:35 -07:00
|
|
|
for key in copy_keys:
|
2015-01-28 19:34:35 -07:00
|
|
|
if key in d["unsigned"]:
|
|
|
|
d[key] = d["unsigned"][key]
|
|
|
|
|
|
|
|
return d
|
|
|
|
|
|
|
|
|
2021-10-13 05:24:07 -06:00
|
|
|
def format_event_for_client_v2(d: JsonDict) -> JsonDict:
|
2015-01-28 19:34:35 -07:00
|
|
|
drop_keys = (
|
2015-01-28 19:45:33 -07:00
|
|
|
"auth_events",
|
|
|
|
"prev_events",
|
|
|
|
"hashes",
|
|
|
|
"signatures",
|
|
|
|
"depth",
|
|
|
|
"origin",
|
|
|
|
"prev_state",
|
2015-01-28 19:34:35 -07:00
|
|
|
)
|
|
|
|
for key in drop_keys:
|
|
|
|
d.pop(key, None)
|
|
|
|
return d
|
|
|
|
|
|
|
|
|
2021-10-13 05:24:07 -06:00
|
|
|
def format_event_for_client_v2_without_room_id(d: JsonDict) -> JsonDict:
|
2015-01-28 19:34:35 -07:00
|
|
|
d = format_event_for_client_v2(d)
|
|
|
|
d.pop("room_id", None)
|
|
|
|
return d
|
|
|
|
|
|
|
|
|
|
|
|
def serialize_event(
|
2021-10-13 05:24:07 -06:00
|
|
|
e: Union[JsonDict, EventBase],
|
|
|
|
time_now_ms: int,
|
2021-12-06 08:51:15 -07:00
|
|
|
*,
|
2021-10-13 05:24:07 -06:00
|
|
|
as_client_event: bool = True,
|
|
|
|
event_format: Callable[[JsonDict], JsonDict] = format_event_for_client_v1,
|
|
|
|
token_id: Optional[str] = None,
|
|
|
|
only_event_fields: Optional[List[str]] = None,
|
|
|
|
include_stripped_room_state: bool = False,
|
|
|
|
) -> JsonDict:
|
2017-04-26 09:18:08 -06:00
|
|
|
"""Serialize event for clients
|
|
|
|
|
|
|
|
Args:
|
2021-10-13 05:24:07 -06:00
|
|
|
e
|
|
|
|
time_now_ms
|
|
|
|
as_client_event
|
2017-04-26 09:18:08 -06:00
|
|
|
event_format
|
|
|
|
token_id
|
|
|
|
only_event_fields
|
2021-10-13 05:24:07 -06:00
|
|
|
include_stripped_room_state: Some events can have stripped room state
|
2021-06-09 12:39:51 -06:00
|
|
|
stored in the `unsigned` field. This is required for invite and knock
|
|
|
|
functionality. If this option is False, that state will be removed from the
|
|
|
|
event before it is returned. Otherwise, it will be kept.
|
2017-04-26 09:18:08 -06:00
|
|
|
|
|
|
|
Returns:
|
2021-10-13 05:24:07 -06:00
|
|
|
The serialized event dictionary.
|
2017-04-26 09:18:08 -06:00
|
|
|
"""
|
2019-01-29 10:26:24 -07:00
|
|
|
|
2014-12-05 09:20:48 -07:00
|
|
|
# FIXME(erikj): To handle the case of presence events and the like
|
|
|
|
if not isinstance(e, EventBase):
|
|
|
|
return e
|
|
|
|
|
2015-01-26 09:11:28 -07:00
|
|
|
time_now_ms = int(time_now_ms)
|
|
|
|
|
2014-12-05 09:20:48 -07:00
|
|
|
# Should this strip out None's?
|
|
|
|
d = {k: v for k, v in e.get_dict().items()}
|
2015-01-08 07:27:04 -07:00
|
|
|
|
2019-01-29 10:26:24 -07:00
|
|
|
d["event_id"] = e.event_id
|
|
|
|
|
2014-12-05 09:20:48 -07:00
|
|
|
if "age_ts" in d["unsigned"]:
|
2015-01-28 19:34:35 -07:00
|
|
|
d["unsigned"]["age"] = time_now_ms - d["unsigned"]["age_ts"]
|
2015-01-28 19:45:33 -07:00
|
|
|
del d["unsigned"]["age_ts"]
|
2014-12-08 02:08:26 -07:00
|
|
|
|
2014-12-11 06:25:19 -07:00
|
|
|
if "redacted_because" in e.unsigned:
|
2015-01-28 19:34:35 -07:00
|
|
|
d["unsigned"]["redacted_because"] = serialize_event(
|
2015-10-30 05:15:37 -06:00
|
|
|
e.unsigned["redacted_because"], time_now_ms, event_format=event_format
|
2014-12-11 06:25:19 -07:00
|
|
|
)
|
|
|
|
|
2015-01-28 19:34:35 -07:00
|
|
|
if token_id is not None:
|
2015-01-28 19:45:33 -07:00
|
|
|
if token_id == getattr(e.internal_metadata, "token_id", None):
|
|
|
|
txn_id = getattr(e.internal_metadata, "txn_id", None)
|
2015-01-28 19:34:35 -07:00
|
|
|
if txn_id is not None:
|
|
|
|
d["unsigned"]["transaction_id"] = txn_id
|
2014-12-11 06:25:19 -07:00
|
|
|
|
2021-06-09 12:39:51 -06:00
|
|
|
# invite_room_state and knock_room_state are a list of stripped room state events
|
|
|
|
# that are meant to provide metadata about a room to an invitee/knocker. They are
|
|
|
|
# intended to only be included in specific circumstances, such as down sync, and
|
|
|
|
# should not be included in any other case.
|
|
|
|
if not include_stripped_room_state:
|
2017-04-26 09:23:30 -06:00
|
|
|
d["unsigned"].pop("invite_room_state", None)
|
2021-06-09 12:39:51 -06:00
|
|
|
d["unsigned"].pop("knock_room_state", None)
|
2017-04-26 09:23:30 -06:00
|
|
|
|
2015-01-28 19:34:35 -07:00
|
|
|
if as_client_event:
|
2016-11-21 10:42:16 -07:00
|
|
|
d = event_format(d)
|
|
|
|
|
2016-11-22 06:42:11 -07:00
|
|
|
if only_event_fields:
|
|
|
|
if not isinstance(only_event_fields, list) or not all(
|
2020-06-16 06:51:47 -06:00
|
|
|
isinstance(f, str) for f in only_event_fields
|
2018-04-15 13:43:35 -06:00
|
|
|
):
|
2016-11-22 06:42:11 -07:00
|
|
|
raise TypeError("only_event_fields must be a list of strings")
|
2016-11-22 02:59:27 -07:00
|
|
|
d = only_fields(d, only_event_fields)
|
2016-11-21 10:42:16 -07:00
|
|
|
|
|
|
|
return d
|
2019-05-09 06:21:57 -06:00
|
|
|
|
|
|
|
|
2020-09-04 04:54:56 -06:00
|
|
|
class EventClientSerializer:
|
2019-05-09 06:21:57 -06:00
|
|
|
"""Serializes events that are to be sent to clients.
|
|
|
|
|
|
|
|
This is used for bundling extra information with any events to be sent to
|
|
|
|
clients.
|
|
|
|
"""
|
|
|
|
|
2022-01-07 07:10:46 -07:00
|
|
|
def serialize_event(
|
2021-10-13 05:24:07 -06:00
|
|
|
self,
|
|
|
|
event: Union[JsonDict, EventBase],
|
|
|
|
time_now: int,
|
2021-12-06 08:51:15 -07:00
|
|
|
*,
|
2022-01-26 06:27:04 -07:00
|
|
|
bundle_aggregations: Optional[Dict[str, "BundledAggregations"]] = None,
|
2021-10-13 05:24:07 -06:00
|
|
|
**kwargs: Any,
|
|
|
|
) -> JsonDict:
|
2019-05-09 06:21:57 -06:00
|
|
|
"""Serializes a single event.
|
|
|
|
|
|
|
|
Args:
|
2021-11-23 04:43:56 -07:00
|
|
|
event: The event being serialized.
|
2021-10-13 05:24:07 -06:00
|
|
|
time_now: The current time in milliseconds
|
2021-12-06 08:51:15 -07:00
|
|
|
bundle_aggregations: Whether to include the bundled aggregations for this
|
|
|
|
event. Only applies to non-state events. (State events never include
|
|
|
|
bundled aggregations.)
|
2019-05-09 06:21:57 -06:00
|
|
|
**kwargs: Arguments to pass to `serialize_event`
|
|
|
|
|
|
|
|
Returns:
|
2021-10-13 05:24:07 -06:00
|
|
|
The serialized event
|
2019-05-09 06:21:57 -06:00
|
|
|
"""
|
2019-05-14 09:59:21 -06:00
|
|
|
# To handle the case of presence events and the like
|
|
|
|
if not isinstance(event, EventBase):
|
2019-07-23 07:00:55 -06:00
|
|
|
return event
|
2019-05-14 09:59:21 -06:00
|
|
|
|
2019-05-14 09:59:21 -06:00
|
|
|
serialized_event = serialize_event(event, time_now, **kwargs)
|
2019-05-14 09:59:21 -06:00
|
|
|
|
2021-12-06 08:51:15 -07:00
|
|
|
# Check if there are any bundled aggregations to include with the event.
|
2022-01-07 07:10:46 -07:00
|
|
|
if bundle_aggregations:
|
|
|
|
event_aggregations = bundle_aggregations.get(event.event_id)
|
|
|
|
if event_aggregations:
|
2022-01-21 03:31:31 -07:00
|
|
|
self._inject_bundled_aggregations(
|
2022-01-07 07:10:46 -07:00
|
|
|
event,
|
|
|
|
time_now,
|
|
|
|
bundle_aggregations[event.event_id],
|
|
|
|
serialized_event,
|
|
|
|
)
|
2021-10-21 12:39:16 -06:00
|
|
|
|
2019-07-23 07:00:55 -06:00
|
|
|
return serialized_event
|
2019-05-09 06:21:57 -06:00
|
|
|
|
2022-01-21 03:31:31 -07:00
|
|
|
def _inject_bundled_aggregations(
|
2022-01-07 07:10:46 -07:00
|
|
|
self,
|
|
|
|
event: EventBase,
|
|
|
|
time_now: int,
|
2022-01-26 06:27:04 -07:00
|
|
|
aggregations: "BundledAggregations",
|
2022-01-07 07:10:46 -07:00
|
|
|
serialized_event: JsonDict,
|
2021-11-23 04:43:56 -07:00
|
|
|
) -> None:
|
2021-12-06 08:51:15 -07:00
|
|
|
"""Potentially injects bundled aggregations into the unsigned portion of the serialized event.
|
2021-11-23 04:43:56 -07:00
|
|
|
|
|
|
|
Args:
|
|
|
|
event: The event being serialized.
|
|
|
|
time_now: The current time in milliseconds
|
2022-01-07 07:10:46 -07:00
|
|
|
aggregations: The bundled aggregation to serialize.
|
2021-11-23 04:43:56 -07:00
|
|
|
serialized_event: The serialized event which may be modified.
|
|
|
|
|
|
|
|
"""
|
2022-01-26 06:27:04 -07:00
|
|
|
serialized_aggregations = {}
|
|
|
|
|
|
|
|
if aggregations.annotations:
|
|
|
|
serialized_aggregations[RelationTypes.ANNOTATION] = aggregations.annotations
|
|
|
|
|
|
|
|
if aggregations.references:
|
|
|
|
serialized_aggregations[RelationTypes.REFERENCE] = aggregations.references
|
2021-11-23 04:43:56 -07:00
|
|
|
|
2022-01-26 06:27:04 -07:00
|
|
|
if aggregations.replace:
|
2021-11-23 04:43:56 -07:00
|
|
|
# If there is an edit replace the content, preserving existing
|
|
|
|
# relations.
|
2022-01-26 06:27:04 -07:00
|
|
|
edit = aggregations.replace
|
2021-11-23 04:43:56 -07:00
|
|
|
|
|
|
|
# Ensure we take copies of the edit content, otherwise we risk modifying
|
|
|
|
# the original event.
|
|
|
|
edit_content = edit.content.copy()
|
|
|
|
|
|
|
|
# Unfreeze the event content if necessary, so that we may modify it below
|
|
|
|
edit_content = unfreeze(edit_content)
|
|
|
|
serialized_event["content"] = edit_content.get("m.new_content", {})
|
|
|
|
|
|
|
|
# Check for existing relations
|
|
|
|
relates_to = event.content.get("m.relates_to")
|
|
|
|
if relates_to:
|
|
|
|
# Keep the relations, ensuring we use a dict copy of the original
|
|
|
|
serialized_event["content"]["m.relates_to"] = relates_to.copy()
|
|
|
|
else:
|
|
|
|
serialized_event["content"].pop("m.relates_to", None)
|
|
|
|
|
2022-01-26 06:27:04 -07:00
|
|
|
serialized_aggregations[RelationTypes.REPLACE] = {
|
2021-11-23 04:43:56 -07:00
|
|
|
"event_id": edit.event_id,
|
|
|
|
"origin_server_ts": edit.origin_server_ts,
|
|
|
|
"sender": edit.sender,
|
|
|
|
}
|
|
|
|
|
|
|
|
# If this event is the start of a thread, include a summary of the replies.
|
2022-01-26 06:27:04 -07:00
|
|
|
if aggregations.thread:
|
|
|
|
serialized_aggregations[RelationTypes.THREAD] = {
|
|
|
|
# Don't bundle aggregations as this could recurse forever.
|
|
|
|
"latest_event": self.serialize_event(
|
|
|
|
aggregations.thread.latest_event, time_now, bundle_aggregations=None
|
|
|
|
),
|
|
|
|
"count": aggregations.thread.count,
|
|
|
|
"current_user_participated": aggregations.thread.current_user_participated,
|
|
|
|
}
|
2021-11-23 04:43:56 -07:00
|
|
|
|
2022-01-07 07:10:46 -07:00
|
|
|
# Include the bundled aggregations in the event.
|
2022-01-26 06:27:04 -07:00
|
|
|
if serialized_aggregations:
|
|
|
|
serialized_event["unsigned"].setdefault("m.relations", {}).update(
|
|
|
|
serialized_aggregations
|
|
|
|
)
|
2022-01-07 07:10:46 -07:00
|
|
|
|
|
|
|
def serialize_events(
|
2021-10-13 05:24:07 -06:00
|
|
|
self, events: Iterable[Union[JsonDict, EventBase]], time_now: int, **kwargs: Any
|
|
|
|
) -> List[JsonDict]:
|
2019-05-09 06:21:57 -06:00
|
|
|
"""Serializes multiple events.
|
|
|
|
|
|
|
|
Args:
|
2021-10-13 05:24:07 -06:00
|
|
|
event
|
|
|
|
time_now: The current time in milliseconds
|
2019-05-09 06:21:57 -06:00
|
|
|
**kwargs: Arguments to pass to `serialize_event`
|
|
|
|
|
|
|
|
Returns:
|
2021-10-13 05:24:07 -06:00
|
|
|
The list of serialized events
|
2019-05-09 06:21:57 -06:00
|
|
|
"""
|
2022-01-07 07:10:46 -07:00
|
|
|
return [
|
|
|
|
self.serialize_event(event, time_now=time_now, **kwargs) for event in events
|
|
|
|
]
|
2020-01-28 04:02:55 -07:00
|
|
|
|
|
|
|
|
|
|
|
def copy_power_levels_contents(
|
|
|
|
old_power_levels: Mapping[str, Union[int, Mapping[str, int]]]
|
2021-10-13 05:24:07 -06:00
|
|
|
) -> Dict[str, Union[int, Dict[str, int]]]:
|
2020-01-28 04:02:55 -07:00
|
|
|
"""Copy the content of a power_levels event, unfreezing frozendicts along the way
|
|
|
|
|
|
|
|
Raises:
|
|
|
|
TypeError if the input does not look like a valid power levels event content
|
|
|
|
"""
|
2020-07-20 11:33:04 -06:00
|
|
|
if not isinstance(old_power_levels, collections.abc.Mapping):
|
2020-01-28 04:02:55 -07:00
|
|
|
raise TypeError("Not a valid power-levels content: %r" % (old_power_levels,))
|
|
|
|
|
2021-10-13 05:24:07 -06:00
|
|
|
power_levels: Dict[str, Union[int, Dict[str, int]]] = {}
|
2020-01-28 04:02:55 -07:00
|
|
|
for k, v in old_power_levels.items():
|
|
|
|
|
|
|
|
if isinstance(v, int):
|
|
|
|
power_levels[k] = v
|
|
|
|
continue
|
|
|
|
|
2020-07-20 11:33:04 -06:00
|
|
|
if isinstance(v, collections.abc.Mapping):
|
2021-10-13 05:24:07 -06:00
|
|
|
h: Dict[str, int] = {}
|
|
|
|
power_levels[k] = h
|
2020-01-28 04:02:55 -07:00
|
|
|
for k1, v1 in v.items():
|
|
|
|
# we should only have one level of nesting
|
|
|
|
if not isinstance(v1, int):
|
|
|
|
raise TypeError(
|
2020-01-28 04:08:38 -07:00
|
|
|
"Invalid power_levels value for %s.%s: %r" % (k, k1, v1)
|
2020-01-28 04:02:55 -07:00
|
|
|
)
|
|
|
|
h[k1] = v1
|
|
|
|
continue
|
|
|
|
|
|
|
|
raise TypeError("Invalid power_levels value for %s: %r" % (k, v))
|
|
|
|
|
|
|
|
return power_levels
|
2020-05-14 11:24:01 -06:00
|
|
|
|
|
|
|
|
2021-10-13 05:24:07 -06:00
|
|
|
def validate_canonicaljson(value: Any) -> None:
|
2020-05-14 11:24:01 -06:00
|
|
|
"""
|
|
|
|
Ensure that the JSON object is valid according to the rules of canonical JSON.
|
|
|
|
|
|
|
|
See the appendix section 3.1: Canonical JSON.
|
|
|
|
|
|
|
|
This rejects JSON that has:
|
|
|
|
* An integer outside the range of [-2 ^ 53 + 1, 2 ^ 53 - 1]
|
|
|
|
* Floats
|
|
|
|
* NaN, Infinity, -Infinity
|
|
|
|
"""
|
|
|
|
if isinstance(value, int):
|
2021-08-26 10:07:58 -06:00
|
|
|
if value < CANONICALJSON_MIN_INT or CANONICALJSON_MAX_INT < value:
|
2020-05-14 11:24:01 -06:00
|
|
|
raise SynapseError(400, "JSON integer out of range", Codes.BAD_JSON)
|
|
|
|
|
|
|
|
elif isinstance(value, float):
|
|
|
|
# Note that Infinity, -Infinity, and NaN are also considered floats.
|
|
|
|
raise SynapseError(400, "Bad JSON value: float", Codes.BAD_JSON)
|
|
|
|
|
|
|
|
elif isinstance(value, (dict, frozendict)):
|
|
|
|
for v in value.values():
|
|
|
|
validate_canonicaljson(v)
|
|
|
|
|
|
|
|
elif isinstance(value, (list, tuple)):
|
|
|
|
for i in value:
|
|
|
|
validate_canonicaljson(i)
|
|
|
|
|
|
|
|
elif not isinstance(value, (bool, str)) and value is not None:
|
|
|
|
# Other potential JSON values (bool, None, str) are safe.
|
|
|
|
raise SynapseError(400, "Unknown JSON value", Codes.BAD_JSON)
|