Files
mxaccessgw/clients/python/src/zb_mom_ww_mxgateway/galaxy.py
T
Joseph Doherty a8f86b5336
ci / nightly-windev (push) Has been skipped
ci / java (push) Successful in 2m19s
ci / portable (push) Successful in 23m59s
ci / windows-x86 (push) Failing after 58m42s
test(tst-24): drive the .NET and Python clients against real in-process gRPC servers
TST-24 asked for per-client wire tests against a fake gateway. An audit first
corrected the finding's premise: Go, Rust, and Java already had them — bufconn,
a loopback tonic server, and InProcessServerBuilder respectively — each already
asserting the round trip, the server-observed bearer header, and the ReplayGap
sentinel. The two genuine gaps were .NET (every test substituted the transport
interface; the test project had no server package at all) and Python (stub
monkeypatching everywhere but one opt-in TLS test).

Both now serve mxaccess_gateway.v1.MxAccessGateway over a real transport —
Kestrel h2c and grpc.aio, each on an ephemeral loopback port — and drive the
ordinary public client API against it. Only the gateway's behaviour is canned;
the framing, serialization, metadata, and status codes are genuine. Four shapes
each: full round trip with every reply field asserted, the authorization header
as received by the server (including on the streaming RPC), the ReplayGap
sentinel surfaced as the client's typed signal, and a real PERMISSION_DENIED
mapping to the typed authorization error.

The .NET client was only ever compiled in CI, never tested, so the portable job
gains a dotnet test step.

Fixes a bug the new tests caught on their first run: Python's connect() built the
grpc.aio channel inside asyncio.to_thread, and a grpc.aio channel binds to the
event loop current on the constructing thread, so every non-stub connection
raised 'There is no current event loop in thread'. No mock-based test could see
it, and the test guarding the off-loop behaviour patched create_channel and so
asserted the bug. Split resolve_channel_security (blocking TOFU probe, off-loop)
from create_channel (on-loop); the guard tests now assert both halves.
2026-08-10 08:22:25 -04:00

378 lines
13 KiB
Python

"""Async Galaxy Repository client wrapper.
Wraps the read-only ``GalaxyRepository`` gRPC service exposed by the
MxAccess Gateway. The service lets callers test connectivity to the AVEVA
System Platform Galaxy Repository (ZB SQL database), read the last
deployment timestamp, and enumerate the deployed object hierarchy plus the
attributes on each object.
"""
from __future__ import annotations
import asyncio
from collections.abc import AsyncIterator, Sequence
from datetime import datetime
from typing import Any
import grpc
from google.protobuf.timestamp_pb2 import Timestamp
from .auth import merge_metadata
from .errors import MxGatewayError, map_rpc_error
from .generated import galaxy_repository_pb2 as galaxy_pb
from .generated import galaxy_repository_pb2_grpc as galaxy_pb_grpc
from .options import (
BrowseChildrenOptions,
ClientOptions,
create_channel,
resolve_channel_security,
)
_DISCOVER_HIERARCHY_PAGE_SIZE = 5000
_BROWSE_CHILDREN_PAGE_SIZE = 500
class GalaxyRepositoryClient:
"""Async client for the Galaxy Repository gRPC service."""
def __init__(
self,
*,
options: ClientOptions,
stub: Any,
channel: grpc.aio.Channel | None = None,
) -> None:
"""Initialize the client with resolved options and a gRPC stub."""
self.options = options
self.raw_stub = stub
self._channel = channel
self._closed = False
@classmethod
async def connect(
cls,
options: ClientOptions | None = None,
*,
endpoint: str | None = None,
api_key: str | None = None,
plaintext: bool = False,
ca_file: str | None = None,
require_certificate_validation: bool = False,
server_name_override: str | None = None,
stub: Any | None = None,
) -> "GalaxyRepositoryClient":
"""Create a client with either a real async channel or a supplied fake stub."""
resolved = options or ClientOptions(
endpoint=endpoint or "",
api_key=api_key,
plaintext=plaintext,
ca_file=ca_file,
require_certificate_validation=require_certificate_validation,
server_name_override=server_name_override,
)
if stub is not None:
return cls(options=resolved, stub=stub)
# Resolving security may perform a blocking TLS certificate probe (TOFU
# default); run that off the event loop so connect never freezes it. The
# channel itself must be built on the loop thread — a grpc.aio channel
# binds to the loop current on the constructing thread, and a worker
# thread has none.
security = await asyncio.to_thread(resolve_channel_security, resolved)
channel = create_channel(resolved, security=security)
return cls(
options=resolved,
stub=galaxy_pb_grpc.GalaxyRepositoryStub(channel),
channel=channel,
)
async def __aenter__(self) -> "GalaxyRepositoryClient":
"""Return self to support ``async with`` usage."""
return self
async def __aexit__(self, *_exc_info: object) -> None:
"""Close the client when leaving an ``async with`` block."""
await self.close()
async def close(self) -> None:
"""Close the owned gRPC channel."""
if self._closed:
return
if self._channel is not None:
await self._channel.close()
self._closed = True
async def test_connection(self) -> bool:
"""Return ``True`` when the gateway can reach the Galaxy Repository DB."""
reply = await self._unary(
"test connection",
self.raw_stub.TestConnection,
galaxy_pb.TestConnectionRequest(),
)
return bool(reply.ok)
async def get_last_deploy_time(self) -> datetime | None:
"""Return the last Galaxy deploy timestamp or ``None`` when unset."""
reply = await self._unary(
"get last deploy time",
self.raw_stub.GetLastDeployTime,
galaxy_pb.GetLastDeployTimeRequest(),
)
if not reply.present:
return None
return reply.time_of_last_deploy.ToDatetime()
async def discover_hierarchy(self) -> list[galaxy_pb.GalaxyObject]:
"""Return the deployed Galaxy object hierarchy as raw proto messages."""
objects: list[galaxy_pb.GalaxyObject] = []
seen_page_tokens: set[str] = set()
page_token = ""
while True:
reply = await self._unary(
"discover hierarchy",
self.raw_stub.DiscoverHierarchy,
galaxy_pb.DiscoverHierarchyRequest(
page_size=_DISCOVER_HIERARCHY_PAGE_SIZE,
page_token=page_token,
),
)
objects.extend(reply.objects)
page_token = reply.next_page_token
if not page_token:
return objects
if page_token in seen_page_tokens:
raise MxGatewayError(
f"galaxy discover hierarchy returned repeated page token {page_token!r}"
)
seen_page_tokens.add(page_token)
async def browse_children_raw(
self, request: galaxy_pb.BrowseChildrenRequest
) -> galaxy_pb.BrowseChildrenReply:
"""Issue one BrowseChildren RPC and return the raw reply.
Lower-level escape hatch for callers that need direct page-token control
or do not want LazyBrowseNode wrapping. Most callers should use
:py:meth:`browse` and :py:meth:`LazyBrowseNode.expand` instead.
"""
return await self._unary(
"browse children",
self.raw_stub.BrowseChildren,
request,
)
async def browse(
self,
options: BrowseChildrenOptions | None = None,
) -> list["LazyBrowseNode"]:
"""Return the root browse nodes for lazy hierarchy traversal.
Each returned ``LazyBrowseNode`` wraps a Galaxy object whose direct
children can be loaded on demand by ``await node.expand()``.
"""
effective = options or BrowseChildrenOptions()
return [
node
async for node in self._iter_browse_children(
parent_gobject_id=None,
options=effective,
)
]
async def _iter_browse_children(
self,
*,
parent_gobject_id: int | None,
options: BrowseChildrenOptions,
) -> AsyncIterator["LazyBrowseNode"]:
page_token = ""
seen_page_tokens: set[str] = set()
while True:
request = galaxy_pb.BrowseChildrenRequest(
page_size=_BROWSE_CHILDREN_PAGE_SIZE,
page_token=page_token,
alarm_bearing_only=options.alarm_bearing_only,
historized_only=options.historized_only,
)
if parent_gobject_id is not None:
request.parent_gobject_id = parent_gobject_id
if options.category_ids:
request.category_ids.extend(options.category_ids)
if options.template_chain_contains:
request.template_chain_contains.extend(options.template_chain_contains)
if options.tag_name_glob:
request.tag_name_glob = options.tag_name_glob
if options.include_attributes is not None:
request.include_attributes = options.include_attributes
reply = await self._unary(
"browse children",
self.raw_stub.BrowseChildren,
request,
)
for index, obj in enumerate(reply.children):
hint = (
index < len(reply.child_has_children)
and bool(reply.child_has_children[index])
)
yield LazyBrowseNode(self, obj, hint, options)
page_token = reply.next_page_token
if not page_token:
return
if page_token in seen_page_tokens:
raise MxGatewayError(
f"galaxy browse children returned repeated page token {page_token!r}"
)
seen_page_tokens.add(page_token)
def watch_deploy_events(
self,
last_seen_deploy_time: datetime | None = None,
*,
metadata: Sequence[tuple[str, str]] | None = None,
) -> AsyncIterator[galaxy_pb.DeployEvent]:
"""Stream Galaxy deploy events.
On subscribe the gateway emits the current cached state and then one
event per new deploy time. ``sequence`` is monotonic per server start;
gaps mean events were dropped from the per-subscriber buffer. When
``last_seen_deploy_time`` is supplied and matches the current cached
deploy time the bootstrap event is suppressed.
"""
request = galaxy_pb.WatchDeployEventsRequest()
if last_seen_deploy_time is not None:
timestamp = Timestamp()
timestamp.FromDatetime(last_seen_deploy_time)
request.last_seen_deploy_time.CopyFrom(timestamp)
kwargs: dict[str, Any] = {"metadata": merge_metadata(self.options.api_key, metadata)}
if self.options.stream_timeout is not None:
kwargs["timeout"] = self.options.stream_timeout
try:
call = self.raw_stub.WatchDeployEvents(request, **kwargs)
except TypeError as error:
if "timeout" not in kwargs or "unexpected keyword argument 'timeout'" not in str(error):
raise
kwargs.pop("timeout")
call = self.raw_stub.WatchDeployEvents(request, **kwargs)
return _canceling_iterator(call)
async def _unary(
self,
operation: str,
method: Any,
request: Any,
*,
metadata: Sequence[tuple[str, str]] | None = None,
) -> Any:
kwargs: dict[str, Any] = {"metadata": merge_metadata(self.options.api_key, metadata)}
if self.options.call_timeout is not None:
kwargs["timeout"] = self.options.call_timeout
try:
call = method(request, **kwargs)
except TypeError as error:
if "timeout" not in kwargs or "unexpected keyword argument 'timeout'" not in str(error):
raise
kwargs.pop("timeout")
call = method(request, **kwargs)
try:
return await call
except asyncio.CancelledError:
cancel = getattr(call, "cancel", None)
if cancel is not None:
cancel()
raise
except grpc.RpcError as error:
raise map_rpc_error(operation, error) from error
class LazyBrowseNode:
"""One node in a lazy-loaded Galaxy browse tree.
Calling ``expand`` once fetches direct children (paginating as needed)
and populates ``children``. Subsequent calls are no-ops so callers can
drive UI expand toggles without de-duping.
"""
def __init__(
self,
client: "GalaxyRepositoryClient",
obj: galaxy_pb.GalaxyObject,
has_children_hint: bool,
options: BrowseChildrenOptions,
) -> None:
"""Initialize a node bound to its owning client and filter set."""
self._client = client
self._object = obj
self._has_children_hint = has_children_hint
self._options = options
self._children: list[LazyBrowseNode] = []
self._is_expanded = False
self._expand_lock = asyncio.Lock()
@property
def object(self) -> galaxy_pb.GalaxyObject:
"""Return the underlying ``GalaxyObject`` proto for this node."""
return self._object
@property
def has_children_hint(self) -> bool:
"""Return the server hint about whether this node has children."""
return self._has_children_hint
@property
def children(self) -> list["LazyBrowseNode"]:
"""Return a copy of the loaded child nodes (empty until expanded)."""
return list(self._children)
@property
def is_expanded(self) -> bool:
"""Return whether ``expand`` has already populated ``children``."""
return self._is_expanded
async def expand(self) -> None:
"""Fetch direct children of this node; no-op on subsequent calls."""
if self._is_expanded:
return
async with self._expand_lock:
if self._is_expanded:
return
new_children: list[LazyBrowseNode] = []
async for child in self._client._iter_browse_children(
parent_gobject_id=self._object.gobject_id,
options=self._options,
):
new_children.append(child)
self._children.extend(new_children)
self._is_expanded = True
async def _canceling_iterator(call: Any) -> AsyncIterator[galaxy_pb.DeployEvent]:
try:
async for event in call:
yield event
except asyncio.CancelledError:
cancel = getattr(call, "cancel", None)
if cancel is not None:
cancel()
raise
except grpc.RpcError as error:
raise map_rpc_error("watch deploy events", error) from error
finally:
cancel = getattr(call, "cancel", None)
if cancel is not None:
cancel()