refactor(gateway): migrate Discord adapter to bundled plugin (full Teams parity)

First migration of an existing built-in platform adapter to the plugin
system established by IRC / Teams / LINE / Google Chat. Closes #24325;
advances the umbrella refactor in #3823.

Matches Teams' shape exactly — adapter under ``plugins/platforms/discord/``
with the standard ``__init__.py`` / ``adapter.py`` / ``plugin.yaml``
shell, ``register(ctx)`` entry point, **no back-compat shim** at the old
import path, and full parity for the four hooks Teams uses plus the
``apply_yaml_config_fn`` hook that landed in #25443 (the Discord plugin
is the first consumer of that hook):

* ``standalone_sender_fn`` — out-of-process cron delivery via REST API
* ``setup_fn`` — interactive ``hermes setup gateway`` wizard
* ``apply_yaml_config_fn`` — translate ``config.yaml`` ``discord:`` keys
  into ``DISCORD_*`` env vars (replaces the hardcoded block in
  ``gateway/config.py``)
* ``is_connected`` — declares connection state from ``DISCORD_BOT_TOKEN``
* ``check_fn`` — lazy-installs ``discord.py`` on demand
* plus ``allowed_users_env``, ``allow_all_env``, ``cron_deliver_env_var``,
  ``max_message_length``, ``emoji``, ``required_env``, ``install_hint``

* ``gateway/platforms/discord.py`` (5,101 LOC) →
  ``plugins/platforms/discord/adapter.py`` (git rename, R090).
* New ``plugins/platforms/discord/{__init__.py, plugin.yaml}`` with
  ``requires_env`` / ``optional_env`` declarations.
* Append ``register(ctx)`` block + new hook implementations
  (``_standalone_send``, ``interactive_setup``, ``_apply_yaml_config``,
  ``_clean_discord_user_ids``, ``_is_connected``, ``_build_adapter``,
  plus helpers ``_DISCORD_CHANNEL_TYPE_PROBE_CACHE`` etc.) to the
  adapter.

* Replace the ``Platform.DISCORD elif`` branch in
  ``GatewayRunner._create_adapter()`` (−9 LOC) with a generic post-creation
  hook (+6 LOC) in the registry path: any plugin adapter that declares a
  ``gateway_runner`` attribute now gets it auto-injected. Webhook's
  built-in branch is unchanged (it doesn't go through the registry path).

* Move ``_send_discord`` (190 LOC) and helpers
  (``_DISCORD_CHANNEL_TYPE_PROBE_CACHE``, ``_remember_channel_is_forum``,
  ``_probe_is_forum_cached``, ``_derive_forum_thread_name``) from
  ``tools/send_message_tool.py`` into the plugin as ``_standalone_send``.
* Wire via ``standalone_sender_fn=_standalone_send`` (Teams pattern; same
  gap fixed in #21804 for other plugin platforms).
* Replace the Discord ``elif`` in ``tools/send_message_tool.py``
  ``_send_to_platform`` with a 10-line registry-hook dispatch.
* Drop the ``DiscordAdapter`` import and the
  ``Platform.DISCORD: DiscordAdapter.MAX_MESSAGE_LENGTH`` ``_MAX_LENGTHS``
  entry — the registry's ``max_message_length=2000`` covers it.

* Move ``_setup_discord`` and ``_clean_discord_user_ids`` (68 LOC) from
  ``hermes_cli/setup.py`` into the plugin as ``interactive_setup``.
* Wire via ``setup_fn=interactive_setup``.  CLI helpers (``prompt``,
  ``print_info``, etc.) are lazy-imported so the plugin's module-load
  surface stays minimal.
* Remove ``"discord": _s._setup_discord`` from
  ``hermes_cli/gateway.py::_builtin_setup_fn``.
* Remove the entire 32-line ``_PLATFORMS["discord"]`` static dict entry —
  Discord's setup metadata is now discovered dynamically via
  ``_all_platforms()`` from the registry entry.

* Move the 59-line ``discord_cfg`` YAML→env bridge from
  ``gateway/config.py::load_gateway_config()`` into the plugin as
  ``_apply_yaml_config``.  Covers ``require_mention``,
  ``thread_require_mention``, ``free_response_channels``, ``auto_thread``,
  ``reactions``, ``ignored_channels``, ``allowed_channels``,
  ``no_thread_channels``, ``allow_mentions.{everyone,roles,users,
  replied_user}``, and ``reply_to_mode`` (including the YAML 1.1
  ``off``-as-False coercion and the ``extra.reply_to_mode`` fallback).
* Wire via ``apply_yaml_config_fn=_apply_yaml_config``.
* The hook runs BEFORE ``_apply_env_overrides`` and after the generic
  shared-key loop, exactly as documented in
  ``website/docs/developer-guide/adding-platform-adapters.md``.
* Behavior is preserved exactly — every assignment still uses
  ``not os.getenv(...)`` guards so env vars take precedence over YAML.

All 78 references to the old import path are rewritten — no back-compat
shim:

* 51 ``from gateway.platforms.discord import X`` →
  ``from plugins.platforms.discord.adapter import X``
* 5 ``import gateway.platforms.discord as discord_platform`` →
  ``import plugins.platforms.discord.adapter as discord_platform``
* 1 ``from gateway.platforms import discord as discord_mod`` →
  ``from plugins.platforms.discord import adapter as discord_mod``
* 21 ``mock.patch("gateway.platforms.discord.X")`` strings →
  ``mock.patch("plugins.platforms.discord.adapter.X")``
* 1 docstring reference in ``hermes_cli/commands.py``
* 1 import in ``tools/send_message_tool.py`` (now removed entirely)

The import-safety test in ``tests/gateway/test_discord_imports.py`` is
updated to purge the new canonical module name from ``sys.modules``.

**38 files changed, +621 / −473** — net positive due to the YAML hook
implementation (89 new LOC in the plugin trading for 59 deleted in core),
but every line moved has a clear plugin home now.  The git rename is
detected at R090 because the adapter gained ~340 LOC of moved-in hook
implementations (``_standalone_send`` + ``interactive_setup`` +
``_apply_yaml_config`` + helpers).

* All 568 Discord-specific tests pass across 25 ``test_discord_*.py``
  files plus voice/send/text-batching/reload-skills/stream-consumer/
  integration tests.
* All 147 tests in the YAML-touching subset
  (``test_discord_reply_mode``, ``test_discord_free_response``,
  ``test_discord_allowed_channels``, ``test_discord_allowed_mentions``,
  ``test_discord_channel_controls``, ``test_discord_reactions``,
  ``test_discord_thread_persistence``, ``test_runtime_footer``) pass —
  this is the strongest signal that the YAML→env hook behaves
  identically to the legacy block.
* Broader gateway/cron/integration sweep (1297 tests) introduces zero
  new failures vs ``main``.  Pre-existing failures in
  ``tests/gateway/test_tts_media_routing.py`` and
  ``tests/e2e/test_platform_commands.py`` reproduce identically on the
  unchanged ``main`` revision.
* Plugin discovery sanity check confirms Discord registers alongside the
  other four platform plugins:

    Registered platforms: ['discord', 'google_chat', 'irc', 'line', 'teams']

These Discord-shaped tendrils in core were **deliberately not moved** —
they are generic platform-registry concerns affecting every platform,
not Discord-specific:

* ``gateway/config.py:1205`` ``DISCORD_BOT_TOKEN → config.token`` env
  enablement — same shape Telegram has.  The existing
  ``env_enablement_fn`` registry hook only seeds ``extra``, not
  ``.token``, so it can't replace this without an adapter refactor to
  read from ``extra["bot_token"]``.
* ``gateway/run.py`` voice-mode hooks
  (``self.adapters.get(Platform.DISCORD)`` for
  ``start_voice_mode``/``stop_voice_mode``), role-based auth,
  ``DISCORD_ALLOW_BOTS`` branch in ``_is_user_authorized``,
  ``_UPDATE_ALLOWED_PLATFORMS`` frozenset, and the per-platform
  allowlist maps — generic platform-registry concerns.
* ``Platform.DISCORD`` enum literal — stable identifier used as dict
  keys throughout the codebase; removing it is a separate refactor with
  no real benefit.
* ``tools/discord_tool.py`` and ``tools/environments/local.py`` —
  first-class agent tools and env-passthrough config, neither is the
  gateway adapter.

Each of these is worth its own scoping issue when the time comes.
This commit is contained in:
kshitijk4poor 2026-05-12 16:59:12 +05:30 committed by Teknium
parent 4f988634f8
commit cc8e5ec2af
41 changed files with 728 additions and 504 deletions

View file

@ -563,7 +563,6 @@ async def _send_to_platform(platform, pconfig, chat_id, message, thread_id=None,
"""
from gateway.config import Platform
from gateway.platforms.base import BasePlatformAdapter, utf16_len
from gateway.platforms.discord import DiscordAdapter
from gateway.platforms.slack import SlackAdapter
# Telegram adapter import is optional (requires python-telegram-bot)
@ -589,10 +588,10 @@ async def _send_to_platform(platform, pconfig, chat_id, message, thread_id=None,
except Exception:
logger.debug("Failed to apply Slack mrkdwn formatting in _send_to_platform", exc_info=True)
# Platform message length limits (from adapter class attributes)
# Platform message length limits (from adapter class attributes for
# built-in platforms; from PlatformEntry.max_message_length for plugins).
_MAX_LENGTHS = {
Platform.TELEGRAM: TelegramAdapter.MAX_MESSAGE_LENGTH if _telegram_available else 4096,
Platform.DISCORD: DiscordAdapter.MAX_MESSAGE_LENGTH,
Platform.SLACK: SlackAdapter.MAX_MESSAGE_LENGTH,
}
if _feishu_available:
@ -642,17 +641,27 @@ async def _send_to_platform(platform, pconfig, chat_id, message, thread_id=None,
if platform == Platform.WEIXIN:
return await _send_weixin(pconfig, chat_id, message, media_files=media_files)
# --- Discord: special handling for media attachments ---
# --- Discord: chunked delivery via the registry's standalone_sender_fn.
# The plugin's ``_standalone_send`` (registered in
# plugins/platforms/discord/adapter.py) handles forum channels, threads,
# and multipart media uploads. ``_send_via_adapter`` tries the live
# in-process adapter first via ``adapter.send()``, but Discord's elif
# historically went straight to the HTTP path; we preserve that by
# explicitly invoking the registry hook here so behavior is unchanged.
if platform == Platform.DISCORD:
from gateway.platform_registry import platform_registry
entry = platform_registry.get("discord")
if entry is None or entry.standalone_sender_fn is None:
return {"error": "Discord plugin not registered or missing standalone_sender_fn"}
last_result = None
for i, chunk in enumerate(chunks):
is_last = (i == len(chunks) - 1)
result = await _send_discord(
pconfig.token,
result = await entry.standalone_sender_fn(
pconfig,
chat_id,
chunk,
media_files=media_files if is_last else [],
thread_id=thread_id,
media_files=media_files if is_last else [],
)
if isinstance(result, dict) and result.get("error"):
return result
@ -1026,227 +1035,6 @@ async def _send_telegram(token, chat_id, message, media_files=None, thread_id=No
return _error(f"Telegram send failed: {e}")
def _derive_forum_thread_name(message: str) -> str:
"""Derive a thread name from the first line of the message, capped at 100 chars."""
first_line = message.strip().split("\n", 1)[0].strip()
# Strip common markdown heading prefixes
first_line = first_line.lstrip("#").strip()
if not first_line:
first_line = "New Post"
return first_line[:100]
# Process-local cache for Discord channel-type probes. Avoids re-probing the
# same channel on every send when the directory cache has no entry (e.g. fresh
# install, or channel created after the last directory build).
_DISCORD_CHANNEL_TYPE_PROBE_CACHE: Dict[str, bool] = {}
def _remember_channel_is_forum(chat_id: str, is_forum: bool) -> None:
_DISCORD_CHANNEL_TYPE_PROBE_CACHE[str(chat_id)] = bool(is_forum)
def _probe_is_forum_cached(chat_id: str) -> Optional[bool]:
return _DISCORD_CHANNEL_TYPE_PROBE_CACHE.get(str(chat_id))
async def _send_discord(token, chat_id, message, thread_id=None, media_files=None):
"""Send a single message via Discord REST API (no websocket client needed).
Chunking is handled by _send_to_platform() before this is called.
When thread_id is provided, the message is sent directly to that thread
via the /channels/{thread_id}/messages endpoint.
Media files are uploaded one-by-one via multipart/form-data after the
text message is sent (same pattern as Telegram).
Forum channels (type 15) reject POST /messages a thread post is created
automatically via POST /channels/{id}/threads. Media files are uploaded
as multipart attachments on the starter message of the new thread.
Channel type is resolved from the channel directory first, then a
process-local probe cache, and only as a last resort with a live
GET /channels/{id} probe (whose result is memoized).
"""
try:
import aiohttp
except ImportError:
return {"error": "aiohttp not installed. Run: pip install aiohttp"}
try:
from gateway.platforms.base import resolve_proxy_url, proxy_kwargs_for_aiohttp
_proxy = resolve_proxy_url(platform_env_var="DISCORD_PROXY")
_sess_kw, _req_kw = proxy_kwargs_for_aiohttp(_proxy)
auth_headers = {"Authorization": f"Bot {token}"}
json_headers = {**auth_headers, "Content-Type": "application/json"}
media_files = media_files or []
last_data = None
warnings = []
# Thread endpoint: Discord threads are channels; send directly to the thread ID.
if thread_id:
url = f"https://discord.com/api/v10/channels/{thread_id}/messages"
else:
# Check if the target channel is a forum channel (type 15).
# Forum channels reject POST /messages — create a thread post instead.
# Three-layer detection: directory cache → process-local probe
# cache → GET /channels/{id} probe (with result memoized).
_channel_type = None
try:
from gateway.channel_directory import lookup_channel_type
_channel_type = lookup_channel_type("discord", chat_id)
except Exception:
pass
if _channel_type == "forum":
is_forum = True
elif _channel_type is not None:
is_forum = False
else:
cached = _probe_is_forum_cached(chat_id)
if cached is not None:
is_forum = cached
else:
is_forum = False
try:
info_url = f"https://discord.com/api/v10/channels/{chat_id}"
async with aiohttp.ClientSession(timeout=aiohttp.ClientTimeout(total=15), **_sess_kw) as info_sess:
async with info_sess.get(info_url, headers=json_headers, **_req_kw) as info_resp:
if info_resp.status == 200:
info = await info_resp.json()
is_forum = info.get("type") == 15
_remember_channel_is_forum(chat_id, is_forum)
except Exception:
logger.debug("Failed to probe channel type for %s", chat_id, exc_info=True)
if is_forum:
thread_name = _derive_forum_thread_name(message)
thread_url = f"https://discord.com/api/v10/channels/{chat_id}/threads"
# Filter to readable media files up front so we can pick the
# right code path (JSON vs multipart) before opening a session.
valid_media = []
for media_path, _is_voice in media_files:
if not os.path.exists(media_path):
warning = f"Media file not found, skipping: {media_path}"
logger.warning(warning)
warnings.append(warning)
continue
valid_media.append(media_path)
async with aiohttp.ClientSession(timeout=aiohttp.ClientTimeout(total=60), **_sess_kw) as session:
if valid_media:
# Multipart: payload_json + files[N] creates a forum
# thread with the starter message plus attachments in
# a single API call.
attachments_meta = [
{"id": str(idx), "filename": os.path.basename(path)}
for idx, path in enumerate(valid_media)
]
starter_message = {"content": message, "attachments": attachments_meta}
payload_json = json.dumps({"name": thread_name, "message": starter_message})
form = aiohttp.FormData()
form.add_field("payload_json", payload_json, content_type="application/json")
# Buffer file bytes up front — aiohttp's FormData can
# read lazily and we don't want handles closing under
# it on retry.
try:
for idx, media_path in enumerate(valid_media):
with open(media_path, "rb") as fh:
form.add_field(
f"files[{idx}]",
fh.read(),
filename=os.path.basename(media_path),
)
async with session.post(thread_url, headers=auth_headers, data=form, **_req_kw) as resp:
if resp.status not in {200, 201}:
body = await resp.text()
return _error(f"Discord forum thread creation error ({resp.status}): {body}")
data = await resp.json()
except Exception as e:
return _error(_sanitize_error_text(f"Discord forum thread upload failed: {e}"))
else:
# No media — simple JSON POST creates the thread with
# just the text starter.
async with session.post(
thread_url,
headers=json_headers,
json={
"name": thread_name,
"message": {"content": message},
},
**_req_kw,
) as resp:
if resp.status not in {200, 201}:
body = await resp.text()
return _error(f"Discord forum thread creation error ({resp.status}): {body}")
data = await resp.json()
thread_id_created = data.get("id")
starter_msg_id = (data.get("message") or {}).get("id", thread_id_created)
result = {
"success": True,
"platform": "discord",
"chat_id": chat_id,
"thread_id": thread_id_created,
"message_id": starter_msg_id,
}
if warnings:
result["warnings"] = warnings
return result
url = f"https://discord.com/api/v10/channels/{chat_id}/messages"
async with aiohttp.ClientSession(timeout=aiohttp.ClientTimeout(total=30), **_sess_kw) as session:
# Send text message (skip if empty and media is present)
if message.strip() or not media_files:
async with session.post(url, headers=json_headers, json={"content": message}, **_req_kw) as resp:
if resp.status not in {200, 201}:
body = await resp.text()
return _error(f"Discord API error ({resp.status}): {body}")
last_data = await resp.json()
# Send each media file as a separate multipart upload
for media_path, _is_voice in media_files:
if not os.path.exists(media_path):
warning = f"Media file not found, skipping: {media_path}"
logger.warning(warning)
warnings.append(warning)
continue
try:
form = aiohttp.FormData()
filename = os.path.basename(media_path)
with open(media_path, "rb") as f:
form.add_field("files[0]", f, filename=filename)
async with session.post(url, headers=auth_headers, data=form, **_req_kw) as resp:
if resp.status not in {200, 201}:
body = await resp.text()
warning = _sanitize_error_text(f"Failed to send media {media_path}: Discord API error ({resp.status}): {body}")
logger.error(warning)
warnings.append(warning)
continue
last_data = await resp.json()
except Exception as e:
warning = _sanitize_error_text(f"Failed to send media {media_path}: {e}")
logger.error(warning)
warnings.append(warning)
if last_data is None:
error = "No deliverable text or media remained after processing"
if warnings:
return {"error": error, "warnings": warnings}
return {"error": error}
result = {"success": True, "platform": "discord", "chat_id": chat_id, "message_id": last_data.get("id")}
if warnings:
result["warnings"] = warnings
return result
except Exception as e:
return _error(f"Discord send failed: {e}")
async def _send_slack(token, chat_id, message):
"""Send via Slack Web API."""
try: