hermes-agent/tests/tools/test_telegram_send_message_caption.py
kshitijk4poor 709da844b5 feat(gateway): attach MEDIA: caption to the media bubble on standalone sends
hermes send "MEDIA:/x.png This Caption" now arrives as one native captioned
bubble instead of a separate text message followed by an uncaptioned bubble.

Root cause: the standalone senders (hermes send / cron / send_message tool)
stripped the MEDIA: tag, sent the remaining text as its own message, and
called the media send with no caption -- even though hermes send's help
advertises the captioned form and the bridges/adapters already support a
caption. Signal already captioned correctly.

- tools/send_message_tool.py: new _media_caption_split() chokepoint decides
  caption-vs-separate-body (single captionable non-voice file within the
  platform's message-length cap). Wired into the Telegram, WhatsApp and
  Discord dispatch paths.
- Telegram/WhatsApp/Discord: when the single captioned file is missing, the
  caption text is delivered as a plain message so it is never silently lost.
- Telegram caption send gets a MarkdownV2->plain parse fallback.
- Tests: _media_caption_split unit tests + per-platform caption tests
  (ride, multi-file fallback, voice exclusion, over-limit fallback,
  missing-file text fallback); updated the 3 tests that asserted the old
  text-then-media split.

Closes the gap reported against #58911 (the MEDIA_CAPTION directive PR);
credit to @ferreiraesilva for surfacing the caption behavior.
2026-07-09 15:38:32 +05:30

141 lines
5 KiB
Python

"""Standalone Telegram MEDIA:<path> caption delivery.
When `hermes send --to telegram "MEDIA:/x.png This Caption"` carries a single
captionable file plus short text, the text must ride on the media bubble as the
sendPhoto/sendVideo/sendDocument ``caption`` rather than being posted as a
separate sendMessage beforehand. Longer text (> Telegram's 1024 caption cap)
falls back to a separate message. The ``telegram`` package is stubbed.
"""
from __future__ import annotations
import asyncio
import os
import sys
import tempfile
from types import SimpleNamespace
from unittest.mock import AsyncMock, MagicMock
import pytest
def _install_telegram_mock(monkeypatch: pytest.MonkeyPatch, bot_factory: MagicMock) -> None:
parse_mode = SimpleNamespace(MARKDOWN_V2="MarkdownV2", HTML="HTML")
constants_mod = SimpleNamespace(ParseMode=parse_mode)
_MessageEntity = lambda **_kw: SimpleNamespace(**_kw)
telegram_mod = SimpleNamespace(
Bot=bot_factory,
MessageEntity=_MessageEntity,
constants=constants_mod,
)
monkeypatch.setitem(sys.modules, "telegram", telegram_mod)
monkeypatch.setitem(sys.modules, "telegram.constants", constants_mod)
def _make_bot() -> MagicMock:
bot = MagicMock()
bot.send_message = AsyncMock(return_value=SimpleNamespace(message_id=1))
bot.send_photo = AsyncMock(return_value=SimpleNamespace(message_id=2))
bot.send_video = AsyncMock(return_value=SimpleNamespace(message_id=3))
bot.send_document = AsyncMock(return_value=SimpleNamespace(message_id=4))
return bot
def _no_proxy(monkeypatch: pytest.MonkeyPatch) -> None:
for var in (
"TELEGRAM_PROXY", "HTTPS_PROXY", "https_proxy", "HTTP_PROXY",
"http_proxy", "ALL_PROXY", "all_proxy", "NO_PROXY", "no_proxy",
):
monkeypatch.delenv(var, raising=False)
monkeypatch.setattr("gateway.run._gateway_runner_ref", lambda: None, raising=False)
monkeypatch.setattr(sys, "platform", "linux")
def _tmpfile(suffix: str) -> str:
f = tempfile.NamedTemporaryFile(suffix=suffix, delete=False)
f.write(b"x")
f.close()
return f.name
def test_image_caption_rides_bubble_no_separate_text(monkeypatch: pytest.MonkeyPatch) -> None:
from tools.send_message_tool import _send_telegram
_no_proxy(monkeypatch)
bot = _make_bot()
_install_telegram_mock(monkeypatch, MagicMock(return_value=bot))
img = _tmpfile(".png")
try:
res = asyncio.run(
_send_telegram("tok", "123", "This Caption", media_files=[(img, False)])
)
assert res["success"] is True
# No separate text message; caption rides the photo.
bot.send_message.assert_not_awaited()
bot.send_photo.assert_awaited_once()
assert bot.send_photo.await_args.kwargs.get("caption") == "This Caption"
finally:
os.unlink(img)
def test_video_caption_rides_bubble(monkeypatch: pytest.MonkeyPatch) -> None:
from tools.send_message_tool import _send_telegram
_no_proxy(monkeypatch)
bot = _make_bot()
_install_telegram_mock(monkeypatch, MagicMock(return_value=bot))
vid = _tmpfile(".mp4")
try:
res = asyncio.run(
_send_telegram("tok", "123", "Model unit tour", media_files=[(vid, False)])
)
assert res["success"] is True
bot.send_message.assert_not_awaited()
bot.send_video.assert_awaited_once()
assert bot.send_video.await_args.kwargs.get("caption") == "Model unit tour"
finally:
os.unlink(vid)
def test_long_text_falls_back_to_separate_message(monkeypatch: pytest.MonkeyPatch) -> None:
from tools.send_message_tool import _send_telegram
_no_proxy(monkeypatch)
bot = _make_bot()
_install_telegram_mock(monkeypatch, MagicMock(return_value=bot))
img = _tmpfile(".png")
long_text = "x" * 1100 # over Telegram's 1024 caption cap
try:
res = asyncio.run(
_send_telegram("tok", "123", long_text, media_files=[(img, False)])
)
assert res["success"] is True
# Text too long for a caption — sent as its own message, photo uncaptioned.
bot.send_message.assert_awaited()
bot.send_photo.assert_awaited_once()
assert not bot.send_photo.await_args.kwargs.get("caption")
finally:
os.unlink(img)
def test_multi_file_keeps_separate_text(monkeypatch: pytest.MonkeyPatch) -> None:
from tools.send_message_tool import _send_telegram
_no_proxy(monkeypatch)
bot = _make_bot()
_install_telegram_mock(monkeypatch, MagicMock(return_value=bot))
img = _tmpfile(".png")
img2 = _tmpfile(".jpg")
try:
res = asyncio.run(
_send_telegram("tok", "123", "two pics", media_files=[(img, False), (img2, False)])
)
assert res["success"] is True
# Ambiguous caption→file association: text stays a separate message.
bot.send_message.assert_awaited()
assert bot.send_photo.await_count == 2
for call in bot.send_photo.await_args_list:
assert not call.kwargs.get("caption")
finally:
os.unlink(img)
os.unlink(img2)