Merge branch 'develop' of github.com:matrix-org/synapse into anoa/public_rooms_federate

4 files changed, 489 insertions, 83 deletions

diff --git a/synapse/events/__init__.py b/synapse/events/__init__.py
index 12f1eb0a3e..20c1ab4203 100644
--- a/synapse/events/__init__.py
+++ b/synapse/events/__init__.py
@@ -1,5 +1,6 @@
 # -*- coding: utf-8 -*-
 # Copyright 2014-2016 OpenMarket Ltd
+# Copyright 2019 New Vector Ltd
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -18,6 +19,9 @@ from distutils.util import strtobool
 
 import six
 
+from unpaddedbase64 import encode_base64
+
+from synapse.api.constants import KNOWN_ROOM_VERSIONS, EventFormatVersions, RoomVersions
 from synapse.util.caches import intern_dict
 from synapse.util.frozenutils import freeze
 
@@ -41,8 +45,13 @@ class _EventInternalMetadata(object):
     def is_outlier(self):
         return getattr(self, "outlier", False)
 
-    def is_invite_from_remote(self):
-        return getattr(self, "invite_from_remote", False)
+    def is_out_of_band_membership(self):
+        """Whether this is an out of band membership, like an invite or an invite
+        rejection. This is needed as those events are marked as outliers, but
+        they still need to be processed as if they're new events (e.g. updating
+        invite state in the database, relaying to clients, etc).
+        """
+        return getattr(self, "out_of_band_membership", False)
 
     def get_send_on_behalf_of(self):
         """Whether this server should send the event on behalf of another server.
@@ -53,6 +62,21 @@ class _EventInternalMetadata(object):
         """
         return getattr(self, "send_on_behalf_of", None)
 
+    def need_to_check_redaction(self):
+        """Whether the redaction event needs to be rechecked when fetching
+        from the database.
+
+        Starting in room v3 redaction events are accepted up front, and later
+        checked to see if the redacter and redactee's domains match.
+
+        If the sender of the redaction event is allowed to redact any event
+        due to auth rules, then this will always return false.
+
+        Returns:
+            bool
+        """
+        return getattr(self, "recheck_redaction", False)
+
 
 def _event_dict_property(key):
     # We want to be able to use hasattr with the event dict properties.
@@ -159,8 +183,28 @@ class EventBase(object):
     def keys(self):
         return six.iterkeys(self._event_dict)
 
+    def prev_event_ids(self):
+        """Returns the list of prev event IDs. The order matches the order
+        specified in the event, though there is no meaning to it.
+
+        Returns:
+            list[str]: The list of event IDs of this event's prev_events
+        """
+        return [e for e, _ in self.prev_events]
+
+    def auth_event_ids(self):
+        """Returns the list of auth event IDs. The order matches the order
+        specified in the event, though there is no meaning to it.
+
+        Returns:
+            list[str]: The list of event IDs of this event's auth_events
+        """
+        return [e for e, _ in self.auth_events]
+
 
 class FrozenEvent(EventBase):
+    format_version = EventFormatVersions.V1  # All events of this type are V1
+
     def __init__(self, event_dict, internal_metadata_dict={}, rejected_reason=None):
         event_dict = dict(event_dict)
 
@@ -195,22 +239,136 @@ class FrozenEvent(EventBase):
             rejected_reason=rejected_reason,
         )
 
-    @staticmethod
-    def from_event(event):
-        e = FrozenEvent(
-            event.get_pdu_json()
+    def __str__(self):
+        return self.__repr__()
+
+    def __repr__(self):
+        return "<FrozenEvent event_id='%s', type='%s', state_key='%s'>" % (
+            self.get("event_id", None),
+            self.get("type", None),
+            self.get("state_key", None),
+        )
+
+
+class FrozenEventV2(EventBase):
+    format_version = EventFormatVersions.V2  # All events of this type are V2
+
+    def __init__(self, event_dict, internal_metadata_dict={}, rejected_reason=None):
+        event_dict = dict(event_dict)
+
+        # Signatures is a dict of dicts, and this is faster than doing a
+        # copy.deepcopy
+        signatures = {
+            name: {sig_id: sig for sig_id, sig in sigs.items()}
+            for name, sigs in event_dict.pop("signatures", {}).items()
+        }
+
+        assert "event_id" not in event_dict
+
+        unsigned = dict(event_dict.pop("unsigned", {}))
+
+        # We intern these strings because they turn up a lot (especially when
+        # caching).
+        event_dict = intern_dict(event_dict)
+
+        if USE_FROZEN_DICTS:
+            frozen_dict = freeze(event_dict)
+        else:
+            frozen_dict = event_dict
+
+        self._event_id = None
+        self.type = event_dict["type"]
+        if "state_key" in event_dict:
+            self.state_key = event_dict["state_key"]
+
+        super(FrozenEventV2, self).__init__(
+            frozen_dict,
+            signatures=signatures,
+            unsigned=unsigned,
+            internal_metadata_dict=internal_metadata_dict,
+            rejected_reason=rejected_reason,
         )
 
-        e.internal_metadata = event.internal_metadata
+    @property
+    def event_id(self):
+        # We have to import this here as otherwise we get an import loop which
+        # is hard to break.
+        from synapse.crypto.event_signing import compute_event_reference_hash
+
+        if self._event_id:
+            return self._event_id
+        self._event_id = "$" + encode_base64(compute_event_reference_hash(self)[1])
+        return self._event_id
+
+    def prev_event_ids(self):
+        """Returns the list of prev event IDs. The order matches the order
+        specified in the event, though there is no meaning to it.
+
+        Returns:
+            list[str]: The list of event IDs of this event's prev_events
+        """
+        return self.prev_events
+
+    def auth_event_ids(self):
+        """Returns the list of auth event IDs. The order matches the order
+        specified in the event, though there is no meaning to it.
 
-        return e
+        Returns:
+            list[str]: The list of event IDs of this event's auth_events
+        """
+        return self.auth_events
 
     def __str__(self):
         return self.__repr__()
 
     def __repr__(self):
-        return "<FrozenEvent event_id='%s', type='%s', state_key='%s'>" % (
-            self.get("event_id", None),
+        return "<FrozenEventV2 event_id='%s', type='%s', state_key='%s'>" % (
+            self.event_id,
             self.get("type", None),
             self.get("state_key", None),
         )
+
+
+def room_version_to_event_format(room_version):
+    """Converts a room version string to the event format
+
+    Args:
+        room_version (str)
+
+    Returns:
+        int
+    """
+    if room_version not in KNOWN_ROOM_VERSIONS:
+        # We should have already checked version, so this should not happen
+        raise RuntimeError("Unrecognized room version %s" % (room_version,))
+
+    if room_version in (
+        RoomVersions.V1, RoomVersions.V2, RoomVersions.STATE_V2_TEST,
+    ):
+        return EventFormatVersions.V1
+    elif room_version in (RoomVersions.V3,):
+        return EventFormatVersions.V2
+    else:
+        raise RuntimeError("Unrecognized room version %s" % (room_version,))
+
+
+def event_type_from_format_version(format_version):
+    """Returns the python type to use to construct an Event object for the
+    given event format version.
+
+    Args:
+        format_version (int): The event format version
+
+    Returns:
+        type: A type that can be initialized as per the initializer of
+        `FrozenEvent`
+    """
+
+    if format_version == EventFormatVersions.V1:
+        return FrozenEvent
+    elif format_version == EventFormatVersions.V2:
+        return FrozenEventV2
+    else:
+        raise Exception(
+            "No event format %r" % (format_version,)
+        )
diff --git a/synapse/events/builder.py b/synapse/events/builder.py
index e662eaef10..06e01be918 100644
--- a/synapse/events/builder.py
+++ b/synapse/events/builder.py
@@ -13,63 +13,270 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-import copy
+import attr
 
+from twisted.internet import defer
+
+from synapse.api.constants import (
+    KNOWN_EVENT_FORMAT_VERSIONS,
+    KNOWN_ROOM_VERSIONS,
+    MAX_DEPTH,
+    EventFormatVersions,
+)
+from synapse.crypto.event_signing import add_hashes_and_signatures
 from synapse.types import EventID
 from synapse.util.stringutils import random_string
 
-from . import EventBase, FrozenEvent, _event_dict_property
+from . import (
+    _EventInternalMetadata,
+    event_type_from_format_version,
+    room_version_to_event_format,
+)
+
+
+@attr.s(slots=True, cmp=False, frozen=True)
+class EventBuilder(object):
+    """A format independent event builder used to build up the event content
+    before signing the event.
+
+    (Note that while objects of this class are frozen, the
+    content/unsigned/internal_metadata fields are still mutable)
+
+    Attributes:
+        format_version (int): Event format version
+        room_id (str)
+        type (str)
+        sender (str)
+        content (dict)
+        unsigned (dict)
+        internal_metadata (_EventInternalMetadata)
+
+        _state (StateHandler)
+        _auth (synapse.api.Auth)
+        _store (DataStore)
+        _clock (Clock)
+        _hostname (str): The hostname of the server creating the event
+        _signing_key: The signing key to use to sign the event as the server
+    """
+
+    _state = attr.ib()
+    _auth = attr.ib()
+    _store = attr.ib()
+    _clock = attr.ib()
+    _hostname = attr.ib()
+    _signing_key = attr.ib()
+
+    format_version = attr.ib()
+
+    room_id = attr.ib()
+    type = attr.ib()
+    sender = attr.ib()
+
+    content = attr.ib(default=attr.Factory(dict))
+    unsigned = attr.ib(default=attr.Factory(dict))
+
+    # These only exist on a subset of events, so they raise AttributeError if
+    # someone tries to get them when they don't exist.
+    _state_key = attr.ib(default=None)
+    _redacts = attr.ib(default=None)
 
+    internal_metadata = attr.ib(default=attr.Factory(lambda: _EventInternalMetadata({})))
 
-class EventBuilder(EventBase):
-    def __init__(self, key_values={}, internal_metadata_dict={}):
-        signatures = copy.deepcopy(key_values.pop("signatures", {}))
-        unsigned = copy.deepcopy(key_values.pop("unsigned", {}))
+    @property
+    def state_key(self):
+        if self._state_key is not None:
+            return self._state_key
 
-        super(EventBuilder, self).__init__(
-            key_values,
-            signatures=signatures,
-            unsigned=unsigned,
-            internal_metadata_dict=internal_metadata_dict,
+        raise AttributeError("state_key")
+
+    def is_state(self):
+        return self._state_key is not None
+
+    @defer.inlineCallbacks
+    def build(self, prev_event_ids):
+        """Transform into a fully signed and hashed event
+
+        Args:
+            prev_event_ids (list[str]): The event IDs to use as the prev events
+
+        Returns:
+            Deferred[FrozenEvent]
+        """
+
+        state_ids = yield self._state.get_current_state_ids(
+            self.room_id, prev_event_ids,
+        )
+        auth_ids = yield self._auth.compute_auth_events(
+            self, state_ids,
         )
 
-    event_id = _event_dict_property("event_id")
-    state_key = _event_dict_property("state_key")
-    type = _event_dict_property("type")
+        if self.format_version == EventFormatVersions.V1:
+            auth_events = yield self._store.add_event_hashes(auth_ids)
+            prev_events = yield self._store.add_event_hashes(prev_event_ids)
+        else:
+            auth_events = auth_ids
+            prev_events = prev_event_ids
+
+        old_depth = yield self._store.get_max_depth_of(
+            prev_event_ids,
+        )
+        depth = old_depth + 1
+
+        # we cap depth of generated events, to ensure that they are not
+        # rejected by other servers (and so that they can be persisted in
+        # the db)
+        depth = min(depth, MAX_DEPTH)
+
+        event_dict = {
+            "auth_events": auth_events,
+            "prev_events": prev_events,
+            "type": self.type,
+            "room_id": self.room_id,
+            "sender": self.sender,
+            "content": self.content,
+            "unsigned": self.unsigned,
+            "depth": depth,
+            "prev_state": [],
+        }
+
+        if self.is_state():
+            event_dict["state_key"] = self._state_key
 
-    def build(self):
-        return FrozenEvent.from_event(self)
+        if self._redacts is not None:
+            event_dict["redacts"] = self._redacts
+
+        defer.returnValue(
+            create_local_event_from_event_dict(
+                clock=self._clock,
+                hostname=self._hostname,
+                signing_key=self._signing_key,
+                format_version=self.format_version,
+                event_dict=event_dict,
+                internal_metadata_dict=self.internal_metadata.get_dict(),
+            )
+        )
 
 
 class EventBuilderFactory(object):
-    def __init__(self, clock, hostname):
-        self.clock = clock
-        self.hostname = hostname
+    def __init__(self, hs):
+        self.clock = hs.get_clock()
+        self.hostname = hs.hostname
+        self.signing_key = hs.config.signing_key[0]
+
+        self.store = hs.get_datastore()
+        self.state = hs.get_state_handler()
+        self.auth = hs.get_auth()
+
+    def new(self, room_version, key_values):
+        """Generate an event builder appropriate for the given room version
+
+        Args:
+            room_version (str): Version of the room that we're creating an
+                event builder for
+            key_values (dict): Fields used as the basis of the new event
+
+        Returns:
+            EventBuilder
+        """
+
+        # There's currently only the one event version defined
+        if room_version not in KNOWN_ROOM_VERSIONS:
+            raise Exception(
+                "No event format defined for version %r" % (room_version,)
+            )
+
+        return EventBuilder(
+            store=self.store,
+            state=self.state,
+            auth=self.auth,
+            clock=self.clock,
+            hostname=self.hostname,
+            signing_key=self.signing_key,
+            format_version=room_version_to_event_format(room_version),
+            type=key_values["type"],
+            state_key=key_values.get("state_key"),
+            room_id=key_values["room_id"],
+            sender=key_values["sender"],
+            content=key_values.get("content", {}),
+            unsigned=key_values.get("unsigned", {}),
+            redacts=key_values.get("redacts", None),
+        )
+
+
+def create_local_event_from_event_dict(clock, hostname, signing_key,
+                                       format_version, event_dict,
+                                       internal_metadata_dict=None):
+    """Takes a fully formed event dict, ensuring that fields like `origin`
+    and `origin_server_ts` have correct values for a locally produced event,
+    then signs and hashes it.
+
+    Args:
+        clock (Clock)
+        hostname (str)
+        signing_key
+        format_version (int)
+        event_dict (dict)
+        internal_metadata_dict (dict|None)
+
+    Returns:
+        FrozenEvent
+    """
+
+    # There's currently only the one event version defined
+    if format_version not in KNOWN_EVENT_FORMAT_VERSIONS:
+        raise Exception(
+            "No event format defined for version %r" % (format_version,)
+        )
+
+    if internal_metadata_dict is None:
+        internal_metadata_dict = {}
+
+    time_now = int(clock.time_msec())
+
+    if format_version == EventFormatVersions.V1:
+        event_dict["event_id"] = _create_event_id(clock, hostname)
+
+    event_dict["origin"] = hostname
+    event_dict["origin_server_ts"] = time_now
+
+    event_dict.setdefault("unsigned", {})
+    age = event_dict["unsigned"].pop("age", 0)
+    event_dict["unsigned"].setdefault("age_ts", time_now - age)
+
+    event_dict.setdefault("signatures", {})
+
+    add_hashes_and_signatures(
+        event_dict,
+        hostname,
+        signing_key,
+    )
+    return event_type_from_format_version(format_version)(
+        event_dict, internal_metadata_dict=internal_metadata_dict,
+    )
 
-        self.event_id_count = 0
 
-    def create_event_id(self):
-        i = str(self.event_id_count)
-        self.event_id_count += 1
+# A counter used when generating new event IDs
+_event_id_counter = 0
 
-        local_part = str(int(self.clock.time())) + i + random_string(5)
 
-        e_id = EventID(local_part, self.hostname)
+def _create_event_id(clock, hostname):
+    """Create a new event ID
 
-        return e_id.to_string()
+    Args:
+        clock (Clock)
+        hostname (str): The server name for the event ID
 
-    def new(self, key_values={}):
-        key_values["event_id"] = self.create_event_id()
+    Returns:
+        str
+    """
 
-        time_now = int(self.clock.time_msec())
+    global _event_id_counter
 
-        key_values.setdefault("origin", self.hostname)
-        key_values.setdefault("origin_server_ts", time_now)
+    i = str(_event_id_counter)
+    _event_id_counter += 1
 
-        key_values.setdefault("unsigned", {})
-        age = key_values["unsigned"].pop("age", 0)
-        key_values["unsigned"].setdefault("age_ts", time_now - age)
+    local_part = str(int(clock.time())) + i + random_string(5)
 
-        key_values["signatures"] = {}
+    e_id = EventID(local_part, hostname)
 
-        return EventBuilder(key_values=key_values,)
+    return e_id.to_string()
diff --git a/synapse/events/utils.py b/synapse/events/utils.py
index 652941ca0d..07fccdd8f9 100644
--- a/synapse/events/utils.py
+++ b/synapse/events/utils.py
@@ -38,8 +38,31 @@ def prune_event(event):
     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.
+
+    Args:
+        event (FrozenEvent)
+
+    Returns:
+        FrozenEvent
+    """
+    pruned_event_dict = prune_event_dict(event.get_dict())
+
+    from . import event_type_from_format_version
+    return event_type_from_format_version(event.format_version)(
+        pruned_event_dict, event.internal_metadata.get_dict()
+    )
+
+
+def prune_event_dict(event_dict):
+    """Redacts the event_dict in the same way as `prune_event`, except it
+    operates on dicts rather than event objects
+
+    Args:
+        event_dict (dict)
+
+    Returns:
+        dict: A copy of the pruned event dict
     """
-    event_type = event.type
 
     allowed_keys = [
         "event_id",
@@ -59,13 +82,13 @@ def prune_event(event):
         "membership",
     ]
 
-    event_dict = event.get_dict()
+    event_type = event_dict["type"]
 
     new_content = {}
 
     def add_fields(*fields):
         for field in fields:
-            if field in event.content:
+            if field in event_dict["content"]:
                 new_content[field] = event_dict["content"][field]
 
     if event_type == EventTypes.Member:
@@ -98,17 +121,17 @@ def prune_event(event):
 
     allowed_fields["content"] = new_content
 
-    allowed_fields["unsigned"] = {}
+    unsigned = {}
+    allowed_fields["unsigned"] = unsigned
 
-    if "age_ts" in event.unsigned:
-        allowed_fields["unsigned"]["age_ts"] = event.unsigned["age_ts"]
-    if "replaces_state" in event.unsigned:
-        allowed_fields["unsigned"]["replaces_state"] = event.unsigned["replaces_state"]
+    event_unsigned = event_dict.get("unsigned", {})
 
-    return type(event)(
-        allowed_fields,
-        internal_metadata_dict=event.internal_metadata.get_dict()
-    )
+    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
 
 
 def _copy_field(src, dst, field):
@@ -244,6 +267,7 @@ def serialize_event(e, time_now_ms, as_client_event=True,
     Returns:
         dict
     """
+
     # FIXME(erikj): To handle the case of presence events and the like
     if not isinstance(e, EventBase):
         return e
@@ -253,6 +277,8 @@ def serialize_event(e, time_now_ms, as_client_event=True,
     # Should this strip out None's?
     d = {k: v for k, v in e.get_dict().items()}
 
+    d["event_id"] = e.event_id
+
     if "age_ts" in d["unsigned"]:
         d["unsigned"]["age"] = time_now_ms - d["unsigned"]["age_ts"]
         del d["unsigned"]["age_ts"]
diff --git a/synapse/events/validator.py b/synapse/events/validator.py
index cf184748a1..a072674b02 100644
--- a/synapse/events/validator.py
+++ b/synapse/events/validator.py
@@ -15,23 +15,29 @@
 
 from six import string_types
 
-from synapse.api.constants import EventTypes, Membership
+from synapse.api.constants import EventFormatVersions, EventTypes, Membership
 from synapse.api.errors import SynapseError
 from synapse.types import EventID, RoomID, UserID
 
 
 class EventValidator(object):
+    def validate_new(self, event):
+        """Validates the event has roughly the right format
 
-    def validate(self, event):
-        EventID.from_string(event.event_id)
-        RoomID.from_string(event.room_id)
+        Args:
+            event (FrozenEvent)
+        """
+        self.validate_builder(event)
+
+        if event.format_version == EventFormatVersions.V1:
+            EventID.from_string(event.event_id)
 
         required = [
-            # "auth_events",
+            "auth_events",
             "content",
-            # "hashes",
+            "hashes",
             "origin",
-            # "prev_events",
+            "prev_events",
             "sender",
             "type",
         ]
@@ -41,8 +47,25 @@ class EventValidator(object):
                 raise SynapseError(400, "Event does not have key %s" % (k,))
 
         # Check that the following keys have string values
-        strings = [
+        event_strings = [
             "origin",
+        ]
+
+        for s in event_strings:
+            if not isinstance(getattr(event, s), string_types):
+                raise SynapseError(400, "'%s' not a string type" % (s,))
+
+    def validate_builder(self, event):
+        """Validates that the builder/event has roughly the right format. Only
+        checks values that we expect a proto event to have, rather than all the
+        fields an event would have
+
+        Args:
+            event (EventBuilder|FrozenEvent)
+        """
+
+        strings = [
+            "room_id",
             "sender",
             "type",
         ]
@@ -54,22 +77,7 @@ class EventValidator(object):
             if not isinstance(getattr(event, s), string_types):
                 raise SynapseError(400, "Not '%s' a string type" % (s,))
 
-        if event.type == EventTypes.Member:
-            if "membership" not in event.content:
-                raise SynapseError(400, "Content has not membership key")
-
-            if event.content["membership"] not in Membership.LIST:
-                raise SynapseError(400, "Invalid membership key")
-
-        # Check that the following keys have dictionary values
-        # TODO
-
-        # Check that the following keys have the correct format for DAGs
-        # TODO
-
-    def validate_new(self, event):
-        self.validate(event)
-
+        RoomID.from_string(event.room_id)
         UserID.from_string(event.sender)
 
         if event.type == EventTypes.Message:
@@ -86,9 +94,16 @@ class EventValidator(object):
         elif event.type == EventTypes.Name:
             self._ensure_strings(event.content, ["name"])
 
+        elif event.type == EventTypes.Member:
+            if "membership" not in event.content:
+                raise SynapseError(400, "Content has not membership key")
+
+            if event.content["membership"] not in Membership.LIST:
+                raise SynapseError(400, "Invalid membership key")
+
     def _ensure_strings(self, d, keys):
         for s in keys:
             if s not in d:
                 raise SynapseError(400, "'%s' not in content" % (s,))
             if not isinstance(d[s], string_types):
-                raise SynapseError(400, "Not '%s' a string type" % (s,))
+                raise SynapseError(400, "'%s' not a string type" % (s,))