Telegram has a 4096 character limit per message. This fix: - Splits messages longer than 4000 chars into multiple chunks - Prefers breaking at newline boundaries to preserve formatting - Falls back to space boundaries if no newlines available - Forces split at max length if no good boundaries exist - Adds comprehensive tests for message splitting logic
417 lines
14 KiB
Python
417 lines
14 KiB
Python
"""Tests for Telegram channel implementation."""
|
|
|
|
import pytest
|
|
from unittest.mock import AsyncMock, MagicMock
|
|
|
|
from nanobot.bus.events import OutboundMessage
|
|
from nanobot.bus.queue import MessageBus
|
|
from nanobot.channels.telegram import TelegramChannel, _markdown_to_telegram_html
|
|
from nanobot.config.schema import TelegramConfig
|
|
|
|
|
|
def _make_config() -> TelegramConfig:
|
|
return TelegramConfig(
|
|
enabled=True,
|
|
token="fake-token",
|
|
allow_from=[],
|
|
proxy=None,
|
|
)
|
|
|
|
|
|
class TestMarkdownToTelegramHtml:
|
|
"""Tests for markdown to Telegram HTML conversion."""
|
|
|
|
def test_empty_text(self) -> None:
|
|
assert _markdown_to_telegram_html("") == ""
|
|
|
|
def test_plain_text_passthrough(self) -> None:
|
|
text = "Hello world"
|
|
assert _markdown_to_telegram_html(text) == "Hello world"
|
|
|
|
def test_bold_double_asterisks(self) -> None:
|
|
text = "This is **bold** text"
|
|
assert _markdown_to_telegram_html(text) == "This is <b>bold</b> text"
|
|
|
|
def test_bold_double_underscore(self) -> None:
|
|
text = "This is __bold__ text"
|
|
assert _markdown_to_telegram_html(text) == "This is <b>bold</b> text"
|
|
|
|
def test_italic_underscore(self) -> None:
|
|
text = "This is _italic_ text"
|
|
assert _markdown_to_telegram_html(text) == "This is <i>italic</i> text"
|
|
|
|
def test_italic_not_inside_words(self) -> None:
|
|
text = "some_var_name"
|
|
assert _markdown_to_telegram_html(text) == "some_var_name"
|
|
|
|
def test_strikethrough(self) -> None:
|
|
text = "This is ~~deleted~~ text"
|
|
assert _markdown_to_telegram_html(text) == "This is <s>deleted</s> text"
|
|
|
|
def test_inline_code(self) -> None:
|
|
text = "Use `print()` function"
|
|
result = _markdown_to_telegram_html(text)
|
|
assert "<code>print()</code>" in result
|
|
|
|
def test_inline_code_escapes_html(self) -> None:
|
|
text = "Use `<div>` tag"
|
|
result = _markdown_to_telegram_html(text)
|
|
assert "<code><div></code>" in result
|
|
|
|
def test_code_block(self) -> None:
|
|
text = """Here is code:
|
|
```python
|
|
def hello():
|
|
return "world"
|
|
```
|
|
Done.
|
|
"""
|
|
result = _markdown_to_telegram_html(text)
|
|
assert "<pre><code>" in result
|
|
assert "def hello():" in result
|
|
assert "</code></pre>" in result
|
|
|
|
def test_code_block_escapes_html(self) -> None:
|
|
text = """```
|
|
<div>test</div>
|
|
```"""
|
|
result = _markdown_to_telegram_html(text)
|
|
assert "<div>test</div>" in result
|
|
|
|
def test_headers_stripped(self) -> None:
|
|
text = "# Header 1\n## Header 2\n### Header 3"
|
|
result = _markdown_to_telegram_html(text)
|
|
assert "# Header 1" not in result
|
|
assert "Header 1" in result
|
|
assert "Header 2" in result
|
|
assert "Header 3" in result
|
|
|
|
def test_blockquotes_stripped(self) -> None:
|
|
text = "> This is a quote\nMore text"
|
|
result = _markdown_to_telegram_html(text)
|
|
assert "> " not in result
|
|
assert "This is a quote" in result
|
|
|
|
def test_links_converted(self) -> None:
|
|
text = "Check [this link](https://example.com) out"
|
|
result = _markdown_to_telegram_html(text)
|
|
assert '<a href="https://example.com">this link</a>' in result
|
|
|
|
def test_bullet_list_converted(self) -> None:
|
|
text = "- Item 1\n* Item 2"
|
|
result = _markdown_to_telegram_html(text)
|
|
assert "• Item 1" in result
|
|
assert "• Item 2" in result
|
|
|
|
def test_html_special_chars_escaped(self) -> None:
|
|
text = "5 < 10 and 10 > 5"
|
|
result = _markdown_to_telegram_html(text)
|
|
assert "5 < 10" in result
|
|
assert "10 > 5" in result
|
|
|
|
def test_complex_nested_formatting(self) -> None:
|
|
text = "**Bold _and italic_** and `code`"
|
|
result = _markdown_to_telegram_html(text)
|
|
assert "<b>Bold <i>and italic</i></b>" in result
|
|
assert "<code>code</code>" in result
|
|
|
|
|
|
class TestTelegramChannelSend:
|
|
"""Tests for TelegramChannel.send() method."""
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_send_short_message_single_chunk(self, monkeypatch) -> None:
|
|
"""Short messages are sent as a single message."""
|
|
sent_messages = []
|
|
|
|
class FakeBot:
|
|
async def send_message(self, chat_id, text, parse_mode=None):
|
|
sent_messages.append({"chat_id": chat_id, "text": text, "parse_mode": parse_mode})
|
|
|
|
fake_app = MagicMock()
|
|
fake_app.bot = FakeBot()
|
|
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
channel._app = fake_app
|
|
|
|
await channel.send(OutboundMessage(
|
|
channel="telegram",
|
|
chat_id="123456",
|
|
content="Hello world"
|
|
))
|
|
|
|
assert len(sent_messages) == 1
|
|
assert sent_messages[0]["chat_id"] == 123456
|
|
assert "Hello world" in sent_messages[0]["text"]
|
|
assert sent_messages[0]["parse_mode"] == "HTML"
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_send_long_message_split_into_chunks(self, monkeypatch) -> None:
|
|
"""Long messages exceeding 4000 chars are split."""
|
|
sent_messages = []
|
|
|
|
class FakeBot:
|
|
async def send_message(self, chat_id, text, parse_mode=None):
|
|
sent_messages.append({"chat_id": chat_id, "text": text, "parse_mode": parse_mode})
|
|
|
|
fake_app = MagicMock()
|
|
fake_app.bot = FakeBot()
|
|
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
channel._app = fake_app
|
|
|
|
# Create a message longer than 4000 chars
|
|
long_content = "A" * 1000 + "\n" + "B" * 1000 + "\n" + "C" * 1000 + "\n" + "D" * 1000 + "\n" + "E" * 1000
|
|
|
|
await channel.send(OutboundMessage(
|
|
channel="telegram",
|
|
chat_id="123456",
|
|
content=long_content
|
|
))
|
|
|
|
assert len(sent_messages) == 2 # Should be split into 2 messages
|
|
assert all(m["chat_id"] == 123456 for m in sent_messages)
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_send_splits_at_newline_when_possible(self, monkeypatch) -> None:
|
|
"""Message splitting prefers newline boundaries."""
|
|
sent_messages = []
|
|
|
|
class FakeBot:
|
|
async def send_message(self, chat_id, text, parse_mode=None):
|
|
sent_messages.append({"chat_id": chat_id, "text": text, "parse_mode": parse_mode})
|
|
|
|
fake_app = MagicMock()
|
|
fake_app.bot = FakeBot()
|
|
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
channel._app = fake_app
|
|
|
|
# Create content with clear paragraph breaks
|
|
paragraphs = [f"Paragraph {i}: " + "x" * 100 for i in range(50)]
|
|
content = "\n".join(paragraphs)
|
|
|
|
await channel.send(OutboundMessage(
|
|
channel="telegram",
|
|
chat_id="123456",
|
|
content=content
|
|
))
|
|
|
|
# Each chunk should end with a complete paragraph (no partial lines)
|
|
for msg in sent_messages:
|
|
# Message should not start with whitespace after stripping
|
|
text = msg["text"]
|
|
assert text == text.lstrip()
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_send_falls_back_to_space_boundary(self, monkeypatch) -> None:
|
|
"""When no newline available, split at space boundary."""
|
|
sent_messages = []
|
|
|
|
class FakeBot:
|
|
async def send_message(self, chat_id, text, parse_mode=None):
|
|
sent_messages.append({"chat_id": chat_id, "text": text, "parse_mode": parse_mode})
|
|
|
|
fake_app = MagicMock()
|
|
fake_app.bot = FakeBot()
|
|
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
channel._app = fake_app
|
|
|
|
# Long content without newlines but with spaces
|
|
content = "word " * 2000 # ~10000 chars
|
|
|
|
await channel.send(OutboundMessage(
|
|
channel="telegram",
|
|
chat_id="123456",
|
|
content=content
|
|
))
|
|
|
|
assert len(sent_messages) >= 2
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_send_forces_split_when_no_good_boundary(self, monkeypatch) -> None:
|
|
"""When no newline or space, force split at max length."""
|
|
sent_messages = []
|
|
|
|
class FakeBot:
|
|
async def send_message(self, chat_id, text, parse_mode=None):
|
|
sent_messages.append({"chat_id": chat_id, "text": text, "parse_mode": parse_mode})
|
|
|
|
fake_app = MagicMock()
|
|
fake_app.bot = FakeBot()
|
|
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
channel._app = fake_app
|
|
|
|
# Long content without any spaces or newlines
|
|
content = "A" * 10000
|
|
|
|
await channel.send(OutboundMessage(
|
|
channel="telegram",
|
|
chat_id="123456",
|
|
content=content
|
|
))
|
|
|
|
assert len(sent_messages) >= 2
|
|
# Verify all chunks combined equal original
|
|
combined = "".join(m["text"] for m in sent_messages)
|
|
assert combined == content
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_send_invalid_chat_id_logs_error(self, monkeypatch) -> None:
|
|
"""Invalid chat_id should log error and not send."""
|
|
sent_messages = []
|
|
|
|
class FakeBot:
|
|
async def send_message(self, chat_id, text, parse_mode=None):
|
|
sent_messages.append({"chat_id": chat_id, "text": text})
|
|
|
|
fake_app = MagicMock()
|
|
fake_app.bot = FakeBot()
|
|
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
channel._app = fake_app
|
|
|
|
await channel.send(OutboundMessage(
|
|
channel="telegram",
|
|
chat_id="not-a-number",
|
|
content="Hello"
|
|
))
|
|
|
|
assert len(sent_messages) == 0
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_send_html_parse_error_falls_back_to_plain_text(self, monkeypatch) -> None:
|
|
"""When HTML parsing fails, fall back to plain text."""
|
|
sent_messages = []
|
|
call_count = 0
|
|
|
|
class FakeBot:
|
|
async def send_message(self, chat_id, text, parse_mode=None):
|
|
nonlocal call_count
|
|
call_count += 1
|
|
if parse_mode == "HTML" and call_count == 1:
|
|
raise Exception("Bad markup")
|
|
sent_messages.append({"chat_id": chat_id, "text": text, "parse_mode": parse_mode})
|
|
|
|
fake_app = MagicMock()
|
|
fake_app.bot = FakeBot()
|
|
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
channel._app = fake_app
|
|
|
|
await channel.send(OutboundMessage(
|
|
channel="telegram",
|
|
chat_id="123456",
|
|
content="Hello **world**"
|
|
))
|
|
|
|
# Should have 2 calls: first HTML (fails), second plain text (succeeds)
|
|
assert call_count == 2
|
|
assert len(sent_messages) == 1
|
|
assert sent_messages[0]["parse_mode"] is None # Plain text
|
|
assert "Hello **world**" in sent_messages[0]["text"]
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_send_not_running_warns(self, monkeypatch) -> None:
|
|
"""If bot not running, log warning."""
|
|
warning_logged = []
|
|
|
|
def mock_warning(msg, *args):
|
|
warning_logged.append(msg)
|
|
|
|
monkeypatch.setattr("nanobot.channels.telegram.logger", MagicMock(warning=mock_warning))
|
|
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
channel._app = None # Not running
|
|
|
|
await channel.send(OutboundMessage(
|
|
channel="telegram",
|
|
chat_id="123456",
|
|
content="Hello"
|
|
))
|
|
|
|
assert any("not running" in str(m) for m in warning_logged)
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_send_stops_typing_indicator(self, monkeypatch) -> None:
|
|
"""Sending message should stop typing indicator."""
|
|
stopped_chats = []
|
|
|
|
class FakeBot:
|
|
async def send_message(self, chat_id, text, parse_mode=None):
|
|
pass
|
|
|
|
fake_app = MagicMock()
|
|
fake_app.bot = FakeBot()
|
|
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
channel._app = fake_app
|
|
channel._stop_typing = lambda chat_id: stopped_chats.append(chat_id)
|
|
|
|
await channel.send(OutboundMessage(
|
|
channel="telegram",
|
|
chat_id="123456",
|
|
content="Hello"
|
|
))
|
|
|
|
assert "123456" in stopped_chats
|
|
|
|
|
|
class TestTelegramChannelTyping:
|
|
"""Tests for typing indicator functionality."""
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_start_typing_creates_task(self) -> None:
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
|
|
# Mock _typing_loop to avoid actual async execution
|
|
channel._typing_loop = AsyncMock()
|
|
|
|
channel._start_typing("123456")
|
|
|
|
assert "123456" in channel._typing_tasks
|
|
assert not channel._typing_tasks["123456"].done()
|
|
|
|
# Clean up
|
|
channel._stop_typing("123456")
|
|
|
|
def test_stop_typing_cancels_task(self) -> None:
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
|
|
# Create a mock task
|
|
mock_task = MagicMock()
|
|
mock_task.done.return_value = False
|
|
channel._typing_tasks["123456"] = mock_task
|
|
|
|
channel._stop_typing("123456")
|
|
|
|
mock_task.cancel.assert_called_once()
|
|
assert "123456" not in channel._typing_tasks
|
|
|
|
|
|
class TestTelegramChannelMediaExtensions:
|
|
"""Tests for media file extension detection."""
|
|
|
|
def test_get_extension_from_mime_type(self) -> None:
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
|
|
assert channel._get_extension("image", "image/jpeg") == ".jpg"
|
|
assert channel._get_extension("image", "image/png") == ".png"
|
|
assert channel._get_extension("image", "image/gif") == ".gif"
|
|
assert channel._get_extension("audio", "audio/ogg") == ".ogg"
|
|
assert channel._get_extension("audio", "audio/mpeg") == ".mp3"
|
|
|
|
def test_get_extension_fallback_to_type(self) -> None:
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
|
|
assert channel._get_extension("image", None) == ".jpg"
|
|
assert channel._get_extension("voice", None) == ".ogg"
|
|
assert channel._get_extension("audio", None) == ".mp3"
|
|
|
|
def test_get_extension_unknown_type(self) -> None:
|
|
channel = TelegramChannel(_make_config(), MessageBus())
|
|
|
|
assert channel._get_extension("unknown", None) == ""
|