Add type hints to synapse/storage/databases/main/account_data.py (#11546)

This commit is contained in:
Sean Quah 2021-12-13 16:28:10 +00:00 committed by GitHub
parent e5cdb9e233
commit 6da8591f2e
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
4 changed files with 87 additions and 33 deletions

1
changelog.d/11546.misc Normal file
View file

@ -0,0 +1 @@
Add missing type hints to storage classes.

View file

@ -25,7 +25,6 @@ exclude = (?x)
^(
|synapse/storage/databases/__init__.py
|synapse/storage/databases/main/__init__.py
|synapse/storage/databases/main/account_data.py
|synapse/storage/databases/main/cache.py
|synapse/storage/databases/main/devices.py
|synapse/storage/databases/main/e2e_room_keys.py
@ -181,6 +180,9 @@ disallow_untyped_defs = True
[mypy-synapse.state.*]
disallow_untyped_defs = True
[mypy-synapse.storage.databases.main.account_data]
disallow_untyped_defs = True
[mypy-synapse.storage.databases.main.client_ips]
disallow_untyped_defs = True

View file

@ -14,15 +14,25 @@
# limitations under the License.
import logging
from typing import TYPE_CHECKING, Dict, List, Optional, Set, Tuple
from typing import TYPE_CHECKING, Any, Dict, Iterable, List, Optional, Set, Tuple, cast
from synapse.api.constants import AccountDataTypes
from synapse.replication.slave.storage._slaved_id_tracker import SlavedIdTracker
from synapse.replication.tcp.streams import AccountDataStream, TagAccountDataStream
from synapse.storage._base import SQLBaseStore, db_to_json
from synapse.storage.database import DatabasePool
from synapse.storage._base import db_to_json
from synapse.storage.database import (
DatabasePool,
LoggingDatabaseConnection,
LoggingTransaction,
)
from synapse.storage.databases.main.cache import CacheInvalidationWorkerStore
from synapse.storage.engines import PostgresEngine
from synapse.storage.util.id_generators import MultiWriterIdGenerator, StreamIdGenerator
from synapse.storage.util.id_generators import (
AbstractStreamIdGenerator,
AbstractStreamIdTracker,
MultiWriterIdGenerator,
StreamIdGenerator,
)
from synapse.types import JsonDict
from synapse.util import json_encoder
from synapse.util.caches.descriptors import cached
@ -34,13 +44,19 @@ if TYPE_CHECKING:
logger = logging.getLogger(__name__)
class AccountDataWorkerStore(SQLBaseStore):
"""This is an abstract base class where subclasses must implement
`get_max_account_data_stream_id` which can be called in the initializer.
"""
class AccountDataWorkerStore(CacheInvalidationWorkerStore):
def __init__(
self,
database: DatabasePool,
db_conn: LoggingDatabaseConnection,
hs: "HomeServer",
):
super().__init__(database, db_conn, hs)
def __init__(self, database: DatabasePool, db_conn, hs: "HomeServer"):
self._instance_name = hs.get_instance_name()
# `_can_write_to_account_data` indicates whether the current worker is allowed
# to write account data. A value of `True` implies that `_account_data_id_gen`
# is an `AbstractStreamIdGenerator` and not just a tracker.
self._account_data_id_gen: AbstractStreamIdTracker
if isinstance(database.engine, PostgresEngine):
self._can_write_to_account_data = (
@ -61,8 +77,6 @@ class AccountDataWorkerStore(SQLBaseStore):
writers=hs.config.worker.writers.account_data,
)
else:
self._can_write_to_account_data = True
# We shouldn't be running in worker mode with SQLite, but its useful
# to support it for unit tests.
#
@ -70,7 +84,8 @@ class AccountDataWorkerStore(SQLBaseStore):
# `StreamIdGenerator`, otherwise we use `SlavedIdTracker` which gets
# updated over replication. (Multiple writers are not supported for
# SQLite).
if hs.get_instance_name() in hs.config.worker.writers.account_data:
if self._instance_name in hs.config.worker.writers.account_data:
self._can_write_to_account_data = True
self._account_data_id_gen = StreamIdGenerator(
db_conn,
"room_account_data",
@ -90,8 +105,6 @@ class AccountDataWorkerStore(SQLBaseStore):
"AccountDataAndTagsChangeCache", account_max
)
super().__init__(database, db_conn, hs)
def get_max_account_data_stream_id(self) -> int:
"""Get the current max stream ID for account data stream
@ -113,7 +126,9 @@ class AccountDataWorkerStore(SQLBaseStore):
room_id string to per room account_data dicts.
"""
def get_account_data_for_user_txn(txn):
def get_account_data_for_user_txn(
txn: LoggingTransaction,
) -> Tuple[Dict[str, JsonDict], Dict[str, Dict[str, JsonDict]]]:
rows = self.db_pool.simple_select_list_txn(
txn,
"account_data",
@ -132,7 +147,7 @@ class AccountDataWorkerStore(SQLBaseStore):
["room_id", "account_data_type", "content"],
)
by_room = {}
by_room: Dict[str, Dict[str, JsonDict]] = {}
for row in rows:
room_data = by_room.setdefault(row["room_id"], {})
room_data[row["account_data_type"]] = db_to_json(row["content"])
@ -177,7 +192,9 @@ class AccountDataWorkerStore(SQLBaseStore):
A dict of the room account_data
"""
def get_account_data_for_room_txn(txn):
def get_account_data_for_room_txn(
txn: LoggingTransaction,
) -> Dict[str, JsonDict]:
rows = self.db_pool.simple_select_list_txn(
txn,
"room_account_data",
@ -207,7 +224,9 @@ class AccountDataWorkerStore(SQLBaseStore):
The room account_data for that type, or None if there isn't any set.
"""
def get_account_data_for_room_and_type_txn(txn):
def get_account_data_for_room_and_type_txn(
txn: LoggingTransaction,
) -> Optional[JsonDict]:
content_json = self.db_pool.simple_select_one_onecol_txn(
txn,
table="room_account_data",
@ -243,14 +262,16 @@ class AccountDataWorkerStore(SQLBaseStore):
if last_id == current_id:
return []
def get_updated_global_account_data_txn(txn):
def get_updated_global_account_data_txn(
txn: LoggingTransaction,
) -> List[Tuple[int, str, str]]:
sql = (
"SELECT stream_id, user_id, account_data_type"
" FROM account_data WHERE ? < stream_id AND stream_id <= ?"
" ORDER BY stream_id ASC LIMIT ?"
)
txn.execute(sql, (last_id, current_id, limit))
return txn.fetchall()
return cast(List[Tuple[int, str, str]], txn.fetchall())
return await self.db_pool.runInteraction(
"get_updated_global_account_data", get_updated_global_account_data_txn
@ -273,14 +294,16 @@ class AccountDataWorkerStore(SQLBaseStore):
if last_id == current_id:
return []
def get_updated_room_account_data_txn(txn):
def get_updated_room_account_data_txn(
txn: LoggingTransaction,
) -> List[Tuple[int, str, str, str]]:
sql = (
"SELECT stream_id, user_id, room_id, account_data_type"
" FROM room_account_data WHERE ? < stream_id AND stream_id <= ?"
" ORDER BY stream_id ASC LIMIT ?"
)
txn.execute(sql, (last_id, current_id, limit))
return txn.fetchall()
return cast(List[Tuple[int, str, str, str]], txn.fetchall())
return await self.db_pool.runInteraction(
"get_updated_room_account_data", get_updated_room_account_data_txn
@ -299,7 +322,9 @@ class AccountDataWorkerStore(SQLBaseStore):
mapping from room_id string to per room account_data dicts.
"""
def get_updated_account_data_for_user_txn(txn):
def get_updated_account_data_for_user_txn(
txn: LoggingTransaction,
) -> Tuple[Dict[str, JsonDict], Dict[str, Dict[str, JsonDict]]]:
sql = (
"SELECT account_data_type, content FROM account_data"
" WHERE user_id = ? AND stream_id > ?"
@ -316,7 +341,7 @@ class AccountDataWorkerStore(SQLBaseStore):
txn.execute(sql, (user_id, stream_id))
account_data_by_room = {}
account_data_by_room: Dict[str, Dict[str, JsonDict]] = {}
for row in txn:
room_account_data = account_data_by_room.setdefault(row[0], {})
room_account_data[row[1]] = db_to_json(row[2])
@ -353,12 +378,15 @@ class AccountDataWorkerStore(SQLBaseStore):
)
)
def process_replication_rows(self, stream_name, instance_name, token, rows):
def process_replication_rows(
self,
stream_name: str,
instance_name: str,
token: int,
rows: Iterable[Any],
) -> None:
if stream_name == TagAccountDataStream.NAME:
self._account_data_id_gen.advance(instance_name, token)
for row in rows:
self.get_tags_for_user.invalidate((row.user_id,))
self._account_data_stream_cache.entity_has_changed(row.user_id, token)
elif stream_name == AccountDataStream.NAME:
self._account_data_id_gen.advance(instance_name, token)
for row in rows:
@ -372,7 +400,8 @@ class AccountDataWorkerStore(SQLBaseStore):
(row.user_id, row.room_id, row.data_type)
)
self._account_data_stream_cache.entity_has_changed(row.user_id, token)
return super().process_replication_rows(stream_name, instance_name, token, rows)
super().process_replication_rows(stream_name, instance_name, token, rows)
async def add_account_data_to_room(
self, user_id: str, room_id: str, account_data_type: str, content: JsonDict
@ -389,6 +418,7 @@ class AccountDataWorkerStore(SQLBaseStore):
The maximum stream ID.
"""
assert self._can_write_to_account_data
assert isinstance(self._account_data_id_gen, AbstractStreamIdGenerator)
content_json = json_encoder.encode(content)
@ -431,6 +461,7 @@ class AccountDataWorkerStore(SQLBaseStore):
The maximum stream ID.
"""
assert self._can_write_to_account_data
assert isinstance(self._account_data_id_gen, AbstractStreamIdGenerator)
async with self._account_data_id_gen.get_next() as next_id:
await self.db_pool.runInteraction(
@ -452,7 +483,7 @@ class AccountDataWorkerStore(SQLBaseStore):
def _add_account_data_for_user(
self,
txn,
txn: LoggingTransaction,
next_id: int,
user_id: str,
account_data_type: str,

View file

@ -15,11 +15,13 @@
# limitations under the License.
import logging
from typing import Dict, List, Tuple, cast
from typing import Any, Dict, Iterable, List, Tuple, cast
from synapse.replication.tcp.streams import TagAccountDataStream
from synapse.storage._base import db_to_json
from synapse.storage.database import LoggingTransaction
from synapse.storage.databases.main.account_data import AccountDataWorkerStore
from synapse.storage.util.id_generators import AbstractStreamIdGenerator
from synapse.types import JsonDict
from synapse.util import json_encoder
from synapse.util.caches.descriptors import cached
@ -204,6 +206,7 @@ class TagsWorkerStore(AccountDataWorkerStore):
The next account data ID.
"""
assert self._can_write_to_account_data
assert isinstance(self._account_data_id_gen, AbstractStreamIdGenerator)
content_json = json_encoder.encode(content)
@ -230,6 +233,7 @@ class TagsWorkerStore(AccountDataWorkerStore):
The next account data ID.
"""
assert self._can_write_to_account_data
assert isinstance(self._account_data_id_gen, AbstractStreamIdGenerator)
def remove_tag_txn(txn: LoggingTransaction, next_id: int) -> None:
sql = (
@ -258,6 +262,7 @@ class TagsWorkerStore(AccountDataWorkerStore):
next_id: The the revision to advance to.
"""
assert self._can_write_to_account_data
assert isinstance(self._account_data_id_gen, AbstractStreamIdGenerator)
txn.call_after(
self._account_data_stream_cache.entity_has_changed, user_id, next_id
@ -287,6 +292,21 @@ class TagsWorkerStore(AccountDataWorkerStore):
# than the id that the client has.
pass
def process_replication_rows(
self,
stream_name: str,
instance_name: str,
token: int,
rows: Iterable[Any],
) -> None:
if stream_name == TagAccountDataStream.NAME:
self._account_data_id_gen.advance(instance_name, token)
for row in rows:
self.get_tags_for_user.invalidate((row.user_id,))
self._account_data_stream_cache.entity_has_changed(row.user_id, token)
super().process_replication_rows(stream_name, instance_name, token, rows)
class TagsStore(TagsWorkerStore):
pass