Hermes Agent UX Improvements
This commit is contained in:
teknium1
parent
b1f55e3ee5
commit
ededaaa874
23 changed files with 945 additions and 1545 deletions
|
|
@ -659,34 +659,90 @@ class BasePlatformAdapter(ABC):
|
|||
|
||||
def truncate_message(self, content: str, max_length: int = 4096) -> List[str]:
|
||||
"""
|
||||
Split a long message into chunks.
|
||||
|
||||
Split a long message into chunks, preserving code block boundaries.
|
||||
|
||||
When a split falls inside a triple-backtick code block, the fence is
|
||||
closed at the end of the current chunk and reopened (with the original
|
||||
language tag) at the start of the next chunk. Multi-chunk responses
|
||||
receive indicators like ``(1/3)``.
|
||||
|
||||
Args:
|
||||
content: The full message content
|
||||
max_length: Maximum length per chunk (platform-specific)
|
||||
|
||||
|
||||
Returns:
|
||||
List of message chunks
|
||||
"""
|
||||
if len(content) <= max_length:
|
||||
return [content]
|
||||
|
||||
chunks = []
|
||||
while content:
|
||||
if len(content) <= max_length:
|
||||
chunks.append(content)
|
||||
|
||||
INDICATOR_RESERVE = 10 # room for " (XX/XX)"
|
||||
FENCE_CLOSE = "\n```"
|
||||
|
||||
chunks: List[str] = []
|
||||
remaining = content
|
||||
# When the previous chunk ended mid-code-block, this holds the
|
||||
# language tag (possibly "") so we can reopen the fence.
|
||||
carry_lang: Optional[str] = None
|
||||
|
||||
while remaining:
|
||||
# If we're continuing a code block from the previous chunk,
|
||||
# prepend a new opening fence with the same language tag.
|
||||
prefix = f"```{carry_lang}\n" if carry_lang is not None else ""
|
||||
|
||||
# How much body text we can fit after accounting for the prefix,
|
||||
# a potential closing fence, and the chunk indicator.
|
||||
headroom = max_length - INDICATOR_RESERVE - len(prefix) - len(FENCE_CLOSE)
|
||||
if headroom < 1:
|
||||
headroom = max_length // 2
|
||||
|
||||
# Everything remaining fits in one final chunk
|
||||
if len(prefix) + len(remaining) <= max_length - INDICATOR_RESERVE:
|
||||
chunks.append(prefix + remaining)
|
||||
break
|
||||
|
||||
# Try to split at a newline
|
||||
split_idx = content.rfind("\n", 0, max_length)
|
||||
if split_idx == -1:
|
||||
# No newline, split at space
|
||||
split_idx = content.rfind(" ", 0, max_length)
|
||||
if split_idx == -1:
|
||||
# No space either, hard split
|
||||
split_idx = max_length
|
||||
|
||||
chunks.append(content[:split_idx])
|
||||
content = content[split_idx:].lstrip()
|
||||
|
||||
|
||||
# Find a natural split point (prefer newlines, then spaces)
|
||||
region = remaining[:headroom]
|
||||
split_at = region.rfind("\n")
|
||||
if split_at < headroom // 2:
|
||||
split_at = region.rfind(" ")
|
||||
if split_at < 1:
|
||||
split_at = headroom
|
||||
|
||||
chunk_body = remaining[:split_at]
|
||||
remaining = remaining[split_at:].lstrip()
|
||||
|
||||
full_chunk = prefix + chunk_body
|
||||
|
||||
# Walk the chunk line-by-line to determine whether we end
|
||||
# inside an open code block.
|
||||
in_code = carry_lang is not None
|
||||
lang = carry_lang or ""
|
||||
for line in full_chunk.split("\n"):
|
||||
stripped = line.strip()
|
||||
if stripped.startswith("```"):
|
||||
if in_code:
|
||||
in_code = False
|
||||
lang = ""
|
||||
else:
|
||||
in_code = True
|
||||
tag = stripped[3:].strip()
|
||||
lang = tag.split()[0] if tag else ""
|
||||
|
||||
if in_code:
|
||||
# Close the orphaned fence so the chunk is valid on its own
|
||||
full_chunk += FENCE_CLOSE
|
||||
carry_lang = lang
|
||||
else:
|
||||
carry_lang = None
|
||||
|
||||
chunks.append(full_chunk)
|
||||
|
||||
# Append chunk indicators when the response spans multiple messages
|
||||
if len(chunks) > 1:
|
||||
total = len(chunks)
|
||||
chunks = [
|
||||
f"{chunk} ({i + 1}/{total})" for i, chunk in enumerate(chunks)
|
||||
]
|
||||
|
||||
return chunks
|
||||
|
|
|
|||
|
|
@ -8,6 +8,7 @@ Uses python-telegram-bot library for:
|
|||
"""
|
||||
|
||||
import asyncio
|
||||
import re
|
||||
from typing import Dict, List, Optional, Any
|
||||
|
||||
try:
|
||||
|
|
@ -49,6 +50,16 @@ def check_telegram_requirements() -> bool:
|
|||
return TELEGRAM_AVAILABLE
|
||||
|
||||
|
||||
# Matches every character that MarkdownV2 requires to be backslash-escaped
|
||||
# when it appears outside a code span or fenced code block.
|
||||
_MDV2_ESCAPE_RE = re.compile(r'([_*\[\]()~`>#\+\-=|{}.!\\])')
|
||||
|
||||
|
||||
def _escape_mdv2(text: str) -> str:
|
||||
"""Escape Telegram MarkdownV2 special characters with a preceding backslash."""
|
||||
return _MDV2_ESCAPE_RE.sub(r'\\\1', text)
|
||||
|
||||
|
||||
class TelegramAdapter(BasePlatformAdapter):
|
||||
"""
|
||||
Telegram bot adapter.
|
||||
|
|
@ -167,7 +178,7 @@ class TelegramAdapter(BasePlatformAdapter):
|
|||
msg = await self._bot.send_message(
|
||||
chat_id=int(chat_id),
|
||||
text=chunk,
|
||||
parse_mode=ParseMode.MARKDOWN,
|
||||
parse_mode=ParseMode.MARKDOWN_V2,
|
||||
reply_to_message_id=int(reply_to) if reply_to and i == 0 else None,
|
||||
message_thread_id=int(thread_id) if thread_id else None,
|
||||
)
|
||||
|
|
@ -297,14 +308,81 @@ class TelegramAdapter(BasePlatformAdapter):
|
|||
|
||||
def format_message(self, content: str) -> str:
|
||||
"""
|
||||
Format message for Telegram.
|
||||
|
||||
Telegram uses a subset of markdown. We'll use the simpler
|
||||
Markdown mode (not MarkdownV2) for compatibility.
|
||||
Convert standard markdown to Telegram MarkdownV2 format.
|
||||
|
||||
Protected regions (code blocks, inline code) are extracted first so
|
||||
their contents are never modified. Standard markdown constructs
|
||||
(headers, bold, italic, links) are translated to MarkdownV2 syntax,
|
||||
and all remaining special characters are escaped.
|
||||
"""
|
||||
# Basic escaping for Telegram Markdown
|
||||
# In Markdown mode (not V2), only certain characters need escaping
|
||||
return content
|
||||
if not content:
|
||||
return content
|
||||
|
||||
placeholders: dict = {}
|
||||
counter = [0]
|
||||
|
||||
def _ph(value: str) -> str:
|
||||
"""Stash *value* behind a placeholder token that survives escaping."""
|
||||
key = f"\x00PH{counter[0]}\x00"
|
||||
counter[0] += 1
|
||||
placeholders[key] = value
|
||||
return key
|
||||
|
||||
text = content
|
||||
|
||||
# 1) Protect fenced code blocks (``` ... ```)
|
||||
text = re.sub(
|
||||
r'(```(?:[^\n]*\n)?[\s\S]*?```)',
|
||||
lambda m: _ph(m.group(0)),
|
||||
text,
|
||||
)
|
||||
|
||||
# 2) Protect inline code (`...`)
|
||||
text = re.sub(r'(`[^`]+`)', lambda m: _ph(m.group(0)), text)
|
||||
|
||||
# 3) Convert markdown links – escape the display text; inside the URL
|
||||
# only ')' and '\' need escaping per the MarkdownV2 spec.
|
||||
def _convert_link(m):
|
||||
display = _escape_mdv2(m.group(1))
|
||||
url = m.group(2).replace('\\', '\\\\').replace(')', '\\)')
|
||||
return _ph(f'[{display}]({url})')
|
||||
|
||||
text = re.sub(r'\[([^\]]+)\]\(([^)]+)\)', _convert_link, text)
|
||||
|
||||
# 4) Convert markdown headers (## Title) → bold *Title*
|
||||
def _convert_header(m):
|
||||
inner = m.group(1).strip()
|
||||
# Strip redundant bold markers that may appear inside a header
|
||||
inner = re.sub(r'\*\*(.+?)\*\*', r'\1', inner)
|
||||
return _ph(f'*{_escape_mdv2(inner)}*')
|
||||
|
||||
text = re.sub(
|
||||
r'^#{1,6}\s+(.+)$', _convert_header, text, flags=re.MULTILINE
|
||||
)
|
||||
|
||||
# 5) Convert bold: **text** → *text* (MarkdownV2 bold)
|
||||
text = re.sub(
|
||||
r'\*\*(.+?)\*\*',
|
||||
lambda m: _ph(f'*{_escape_mdv2(m.group(1))}*'),
|
||||
text,
|
||||
)
|
||||
|
||||
# 6) Convert italic: *text* (single asterisk) → _text_ (MarkdownV2 italic)
|
||||
text = re.sub(
|
||||
r'\*([^*]+)\*',
|
||||
lambda m: _ph(f'_{_escape_mdv2(m.group(1))}_'),
|
||||
text,
|
||||
)
|
||||
|
||||
# 7) Escape remaining special characters in plain text
|
||||
text = _escape_mdv2(text)
|
||||
|
||||
# 8) Restore placeholders in reverse insertion order so that
|
||||
# nested references (a placeholder inside another) resolve correctly.
|
||||
for key in reversed(list(placeholders.keys())):
|
||||
text = text.replace(key, placeholders[key])
|
||||
|
||||
return text
|
||||
|
||||
async def _handle_text_message(self, update: Update, context: ContextTypes.DEFAULT_TYPE) -> None:
|
||||
"""Handle incoming text messages."""
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue