zulip/zerver/tornado/views.py

173 lines
6.5 KiB
Python
Raw Normal View History

import time
from typing import Optional, Sequence
import orjson
from django.http import HttpRequest, HttpResponse
from django.utils.translation import gettext as _
from zerver.decorator import internal_notify_view, process_client
from zerver.lib.exceptions import JsonableError
from zerver.lib.request import REQ, RequestNotes, has_request_variables
from zerver.lib.response import json_success
from zerver.lib.validator import (
check_bool,
check_int,
check_list,
check_string,
to_non_negative_int,
)
from zerver.models import Client, UserProfile, get_client, get_user_profile_by_id
from zerver.tornado.event_queue import fetch_events, get_client_descriptor, process_notification
from zerver.tornado.exceptions import BadEventQueueIdError
@internal_notify_view(True)
def notify(request: HttpRequest) -> HttpResponse:
process_notification(orjson.loads(request.POST["data"]))
return json_success()
@has_request_variables
def cleanup_event_queue(
request: HttpRequest, user_profile: UserProfile, queue_id: str = REQ()
) -> HttpResponse:
client = get_client_descriptor(str(queue_id))
if client is None:
raise BadEventQueueIdError(queue_id)
if user_profile.id != client.user_profile_id:
raise JsonableError(_("You are not authorized to access this queue"))
log_data = RequestNotes.get_notes(request).log_data
assert log_data is not None
log_data["extra"] = f"[{queue_id}]"
client.cleanup()
return json_success()
@internal_notify_view(True)
@has_request_variables
def get_events_internal(
request: HttpRequest, user_profile_id: int = REQ(json_validator=check_int)
) -> HttpResponse:
user_profile = get_user_profile_by_id(user_profile_id)
RequestNotes.get_notes(request).requestor_for_logs = user_profile.format_requestor_for_logs()
process_client(request, user_profile, client_name="internal")
tornado: Rewrite Django integration to duplicate less code. Since essentially the first use of Tornado in Zulip, we've been maintaining our Tornado+Django system, AsyncDjangoHandler, with several hundred lines of Django code copied into it. The goal for that code was simple: We wanted a way to use our Django middleware (for code sharing reasons) inside a Tornado process (since we wanted to use Tornado for our async events system). As part of the Django 2.2.x upgrade, I looked at upgrading this implementation to be based off modern Django, and it's definitely possible to do that: * Continue forking load_middleware to save response middleware. * Continue manually running the Django response middleware. * Continue working out a hack involving copying all of _get_response to change a couple lines allowing us our Tornado code to not actually return the Django HttpResponse so we can long-poll. The previous hack of returning None stopped being viable with the Django 2.2 MiddlewareMixin.__call__ implementation. But I decided to take this opportunity to look at trying to avoid copying material Django code, and there is a way to do it: * Replace RespondAsynchronously with a response.asynchronous attribute on the HttpResponse; this allows Django to run its normal plumbing happily in a way that should be stable over time, and then we proceed to discard the response inside the Tornado `get()` method to implement long-polling. (Better yet might be raising an exception?). This lets us eliminate maintaining a patched copy of _get_response. * Removing the @asynchronous decorator, which didn't add anything now that we only have one API endpoint backend (with two frontend call points) that could call into this. Combined with the last bullet, this lets us remove a significant hack from our never_cache_responses function. * Calling the normal Django `get_response` method from zulip_finish after creating a duplicate request to process, rather than writing totally custom code to do that. This lets us eliminate maintaining a patched copy of Django's load_middleware. * Adding detailed comments explaining how this is supposed to work, what problems we encounter, and how we solve various problems, which is critical to being able to modify this code in the future. A key advantage of these changes is that the exact same code should work on Django 1.11, Django 2.2, and Django 3.x, because we're no longer copying large blocks of core Django code and thus should be much less vulnerable to refactors. There may be a modest performance downside, in that we now run both request and response middleware twice when longpolling (once for the request we discard). We may be able to avoid the expensive part of it, Zulip's own request/response middleware, with a bit of additional custom code to save work for requests where we're planning to discard the response. Profiling will be important to understanding what's worth doing here.
2020-02-06 22:09:10 +01:00
return get_events_backend(request, user_profile)
tornado: Rewrite Django integration to duplicate less code. Since essentially the first use of Tornado in Zulip, we've been maintaining our Tornado+Django system, AsyncDjangoHandler, with several hundred lines of Django code copied into it. The goal for that code was simple: We wanted a way to use our Django middleware (for code sharing reasons) inside a Tornado process (since we wanted to use Tornado for our async events system). As part of the Django 2.2.x upgrade, I looked at upgrading this implementation to be based off modern Django, and it's definitely possible to do that: * Continue forking load_middleware to save response middleware. * Continue manually running the Django response middleware. * Continue working out a hack involving copying all of _get_response to change a couple lines allowing us our Tornado code to not actually return the Django HttpResponse so we can long-poll. The previous hack of returning None stopped being viable with the Django 2.2 MiddlewareMixin.__call__ implementation. But I decided to take this opportunity to look at trying to avoid copying material Django code, and there is a way to do it: * Replace RespondAsynchronously with a response.asynchronous attribute on the HttpResponse; this allows Django to run its normal plumbing happily in a way that should be stable over time, and then we proceed to discard the response inside the Tornado `get()` method to implement long-polling. (Better yet might be raising an exception?). This lets us eliminate maintaining a patched copy of _get_response. * Removing the @asynchronous decorator, which didn't add anything now that we only have one API endpoint backend (with two frontend call points) that could call into this. Combined with the last bullet, this lets us remove a significant hack from our never_cache_responses function. * Calling the normal Django `get_response` method from zulip_finish after creating a duplicate request to process, rather than writing totally custom code to do that. This lets us eliminate maintaining a patched copy of Django's load_middleware. * Adding detailed comments explaining how this is supposed to work, what problems we encounter, and how we solve various problems, which is critical to being able to modify this code in the future. A key advantage of these changes is that the exact same code should work on Django 1.11, Django 2.2, and Django 3.x, because we're no longer copying large blocks of core Django code and thus should be much less vulnerable to refactors. There may be a modest performance downside, in that we now run both request and response middleware twice when longpolling (once for the request we discard). We may be able to avoid the expensive part of it, Zulip's own request/response middleware, with a bit of additional custom code to save work for requests where we're planning to discard the response. Profiling will be important to understanding what's worth doing here.
2020-02-06 22:09:10 +01:00
def get_events(request: HttpRequest, user_profile: UserProfile) -> HttpResponse:
return get_events_backend(request, user_profile)
@has_request_variables
def get_events_backend(
request: HttpRequest,
user_profile: UserProfile,
# user_client is intended only for internal Django=>Tornado requests
# and thus shouldn't be documented for external use.
user_client: Optional[Client] = REQ(
converter=get_client, default=None, intentionally_undocumented=True
),
last_event_id: Optional[int] = REQ(converter=int, default=None),
queue_id: Optional[str] = REQ(default=None),
# apply_markdown, client_gravatar, all_public_streams, and various
# other parameters are only used when registering a new queue via this
# endpoint. This is a feature used primarily by get_events_internal
# and not expected to be used by third-party clients.
apply_markdown: bool = REQ(
default=False, json_validator=check_bool, intentionally_undocumented=True
),
client_gravatar: bool = REQ(
default=False, json_validator=check_bool, intentionally_undocumented=True
),
slim_presence: bool = REQ(
default=False, json_validator=check_bool, intentionally_undocumented=True
),
all_public_streams: bool = REQ(
default=False, json_validator=check_bool, intentionally_undocumented=True
),
event_types: Optional[Sequence[str]] = REQ(
default=None, json_validator=check_list(check_string), intentionally_undocumented=True
),
dont_block: bool = REQ(default=False, json_validator=check_bool),
narrow: Sequence[Sequence[str]] = REQ(
default=[],
json_validator=check_list(check_list(check_string)),
intentionally_undocumented=True,
),
lifespan_secs: int = REQ(
default=0, converter=to_non_negative_int, intentionally_undocumented=True
),
bulk_message_deletion: bool = REQ(
default=False, json_validator=check_bool, intentionally_undocumented=True
),
stream_typing_notifications: bool = REQ(
default=False, json_validator=check_bool, intentionally_undocumented=True
),
user_settings_object: bool = REQ(
default=False, json_validator=check_bool, intentionally_undocumented=True
),
) -> HttpResponse:
if all_public_streams and not user_profile.can_access_public_streams():
raise JsonableError(_("User not authorized for this query"))
tornado: Rewrite Django integration to duplicate less code. Since essentially the first use of Tornado in Zulip, we've been maintaining our Tornado+Django system, AsyncDjangoHandler, with several hundred lines of Django code copied into it. The goal for that code was simple: We wanted a way to use our Django middleware (for code sharing reasons) inside a Tornado process (since we wanted to use Tornado for our async events system). As part of the Django 2.2.x upgrade, I looked at upgrading this implementation to be based off modern Django, and it's definitely possible to do that: * Continue forking load_middleware to save response middleware. * Continue manually running the Django response middleware. * Continue working out a hack involving copying all of _get_response to change a couple lines allowing us our Tornado code to not actually return the Django HttpResponse so we can long-poll. The previous hack of returning None stopped being viable with the Django 2.2 MiddlewareMixin.__call__ implementation. But I decided to take this opportunity to look at trying to avoid copying material Django code, and there is a way to do it: * Replace RespondAsynchronously with a response.asynchronous attribute on the HttpResponse; this allows Django to run its normal plumbing happily in a way that should be stable over time, and then we proceed to discard the response inside the Tornado `get()` method to implement long-polling. (Better yet might be raising an exception?). This lets us eliminate maintaining a patched copy of _get_response. * Removing the @asynchronous decorator, which didn't add anything now that we only have one API endpoint backend (with two frontend call points) that could call into this. Combined with the last bullet, this lets us remove a significant hack from our never_cache_responses function. * Calling the normal Django `get_response` method from zulip_finish after creating a duplicate request to process, rather than writing totally custom code to do that. This lets us eliminate maintaining a patched copy of Django's load_middleware. * Adding detailed comments explaining how this is supposed to work, what problems we encounter, and how we solve various problems, which is critical to being able to modify this code in the future. A key advantage of these changes is that the exact same code should work on Django 1.11, Django 2.2, and Django 3.x, because we're no longer copying large blocks of core Django code and thus should be much less vulnerable to refactors. There may be a modest performance downside, in that we now run both request and response middleware twice when longpolling (once for the request we discard). We may be able to avoid the expensive part of it, Zulip's own request/response middleware, with a bit of additional custom code to save work for requests where we're planning to discard the response. Profiling will be important to understanding what's worth doing here.
2020-02-06 22:09:10 +01:00
# Extract the Tornado handler from the request
tornado_handler = RequestNotes.get_notes(request).tornado_handler
assert tornado_handler is not None
handler = tornado_handler()
assert handler is not None
tornado: Rewrite Django integration to duplicate less code. Since essentially the first use of Tornado in Zulip, we've been maintaining our Tornado+Django system, AsyncDjangoHandler, with several hundred lines of Django code copied into it. The goal for that code was simple: We wanted a way to use our Django middleware (for code sharing reasons) inside a Tornado process (since we wanted to use Tornado for our async events system). As part of the Django 2.2.x upgrade, I looked at upgrading this implementation to be based off modern Django, and it's definitely possible to do that: * Continue forking load_middleware to save response middleware. * Continue manually running the Django response middleware. * Continue working out a hack involving copying all of _get_response to change a couple lines allowing us our Tornado code to not actually return the Django HttpResponse so we can long-poll. The previous hack of returning None stopped being viable with the Django 2.2 MiddlewareMixin.__call__ implementation. But I decided to take this opportunity to look at trying to avoid copying material Django code, and there is a way to do it: * Replace RespondAsynchronously with a response.asynchronous attribute on the HttpResponse; this allows Django to run its normal plumbing happily in a way that should be stable over time, and then we proceed to discard the response inside the Tornado `get()` method to implement long-polling. (Better yet might be raising an exception?). This lets us eliminate maintaining a patched copy of _get_response. * Removing the @asynchronous decorator, which didn't add anything now that we only have one API endpoint backend (with two frontend call points) that could call into this. Combined with the last bullet, this lets us remove a significant hack from our never_cache_responses function. * Calling the normal Django `get_response` method from zulip_finish after creating a duplicate request to process, rather than writing totally custom code to do that. This lets us eliminate maintaining a patched copy of Django's load_middleware. * Adding detailed comments explaining how this is supposed to work, what problems we encounter, and how we solve various problems, which is critical to being able to modify this code in the future. A key advantage of these changes is that the exact same code should work on Django 1.11, Django 2.2, and Django 3.x, because we're no longer copying large blocks of core Django code and thus should be much less vulnerable to refactors. There may be a modest performance downside, in that we now run both request and response middleware twice when longpolling (once for the request we discard). We may be able to avoid the expensive part of it, Zulip's own request/response middleware, with a bit of additional custom code to save work for requests where we're planning to discard the response. Profiling will be important to understanding what's worth doing here.
2020-02-06 22:09:10 +01:00
if user_client is None:
valid_user_client = RequestNotes.get_notes(request).client
assert valid_user_client is not None
else:
valid_user_client = user_client
events_query = dict(
user_profile_id=user_profile.id,
queue_id=queue_id,
last_event_id=last_event_id,
event_types=event_types,
client_type_name=valid_user_client.name,
all_public_streams=all_public_streams,
lifespan_secs=lifespan_secs,
narrow=narrow,
dont_block=dont_block,
handler_id=handler.handler_id,
)
if queue_id is None:
events_query["new_queue_data"] = dict(
user_profile_id=user_profile.id,
realm_id=user_profile.realm_id,
event_types=event_types,
client_type_name=valid_user_client.name,
apply_markdown=apply_markdown,
client_gravatar=client_gravatar,
slim_presence=slim_presence,
all_public_streams=all_public_streams,
queue_timeout=lifespan_secs,
last_connection_time=time.time(),
narrow=narrow,
bulk_message_deletion=bulk_message_deletion,
stream_typing_notifications=stream_typing_notifications,
user_settings_object=user_settings_object,
)
result = fetch_events(events_query)
if "extra_log_data" in result:
log_data = RequestNotes.get_notes(request).log_data
assert log_data is not None
log_data["extra"] = result["extra_log_data"]
if result["type"] == "async":
tornado: Rewrite Django integration to duplicate less code. Since essentially the first use of Tornado in Zulip, we've been maintaining our Tornado+Django system, AsyncDjangoHandler, with several hundred lines of Django code copied into it. The goal for that code was simple: We wanted a way to use our Django middleware (for code sharing reasons) inside a Tornado process (since we wanted to use Tornado for our async events system). As part of the Django 2.2.x upgrade, I looked at upgrading this implementation to be based off modern Django, and it's definitely possible to do that: * Continue forking load_middleware to save response middleware. * Continue manually running the Django response middleware. * Continue working out a hack involving copying all of _get_response to change a couple lines allowing us our Tornado code to not actually return the Django HttpResponse so we can long-poll. The previous hack of returning None stopped being viable with the Django 2.2 MiddlewareMixin.__call__ implementation. But I decided to take this opportunity to look at trying to avoid copying material Django code, and there is a way to do it: * Replace RespondAsynchronously with a response.asynchronous attribute on the HttpResponse; this allows Django to run its normal plumbing happily in a way that should be stable over time, and then we proceed to discard the response inside the Tornado `get()` method to implement long-polling. (Better yet might be raising an exception?). This lets us eliminate maintaining a patched copy of _get_response. * Removing the @asynchronous decorator, which didn't add anything now that we only have one API endpoint backend (with two frontend call points) that could call into this. Combined with the last bullet, this lets us remove a significant hack from our never_cache_responses function. * Calling the normal Django `get_response` method from zulip_finish after creating a duplicate request to process, rather than writing totally custom code to do that. This lets us eliminate maintaining a patched copy of Django's load_middleware. * Adding detailed comments explaining how this is supposed to work, what problems we encounter, and how we solve various problems, which is critical to being able to modify this code in the future. A key advantage of these changes is that the exact same code should work on Django 1.11, Django 2.2, and Django 3.x, because we're no longer copying large blocks of core Django code and thus should be much less vulnerable to refactors. There may be a modest performance downside, in that we now run both request and response middleware twice when longpolling (once for the request we discard). We may be able to avoid the expensive part of it, Zulip's own request/response middleware, with a bit of additional custom code to save work for requests where we're planning to discard the response. Profiling will be important to understanding what's worth doing here.
2020-02-06 22:09:10 +01:00
# Mark this response with .asynchronous; this will result in
# Tornado discarding the response and instead long-polling the
# request. See zulip_finish for more design details.
handler._request = request
tornado: Rewrite Django integration to duplicate less code. Since essentially the first use of Tornado in Zulip, we've been maintaining our Tornado+Django system, AsyncDjangoHandler, with several hundred lines of Django code copied into it. The goal for that code was simple: We wanted a way to use our Django middleware (for code sharing reasons) inside a Tornado process (since we wanted to use Tornado for our async events system). As part of the Django 2.2.x upgrade, I looked at upgrading this implementation to be based off modern Django, and it's definitely possible to do that: * Continue forking load_middleware to save response middleware. * Continue manually running the Django response middleware. * Continue working out a hack involving copying all of _get_response to change a couple lines allowing us our Tornado code to not actually return the Django HttpResponse so we can long-poll. The previous hack of returning None stopped being viable with the Django 2.2 MiddlewareMixin.__call__ implementation. But I decided to take this opportunity to look at trying to avoid copying material Django code, and there is a way to do it: * Replace RespondAsynchronously with a response.asynchronous attribute on the HttpResponse; this allows Django to run its normal plumbing happily in a way that should be stable over time, and then we proceed to discard the response inside the Tornado `get()` method to implement long-polling. (Better yet might be raising an exception?). This lets us eliminate maintaining a patched copy of _get_response. * Removing the @asynchronous decorator, which didn't add anything now that we only have one API endpoint backend (with two frontend call points) that could call into this. Combined with the last bullet, this lets us remove a significant hack from our never_cache_responses function. * Calling the normal Django `get_response` method from zulip_finish after creating a duplicate request to process, rather than writing totally custom code to do that. This lets us eliminate maintaining a patched copy of Django's load_middleware. * Adding detailed comments explaining how this is supposed to work, what problems we encounter, and how we solve various problems, which is critical to being able to modify this code in the future. A key advantage of these changes is that the exact same code should work on Django 1.11, Django 2.2, and Django 3.x, because we're no longer copying large blocks of core Django code and thus should be much less vulnerable to refactors. There may be a modest performance downside, in that we now run both request and response middleware twice when longpolling (once for the request we discard). We may be able to avoid the expensive part of it, Zulip's own request/response middleware, with a bit of additional custom code to save work for requests where we're planning to discard the response. Profiling will be important to understanding what's worth doing here.
2020-02-06 22:09:10 +01:00
response = json_success()
response.asynchronous = True
return response
if result["type"] == "error":
raise result["exception"]
return json_success(result["response"])