vibego 0.2.35__py3-none-any.whl → 0.2.37__py3-none-any.whl

bot.py

@@ -482,6 +482,11 @@ _ESCAPED_CODE_BLOCK_PATTERN = re.compile(
     re.DOTALL
 )
 
+_OPTIMIZE_CHECKBOX_PATTERN = re.compile(r"\\\[\s+\\\]")
+_OPTIMIZE_DOT_PATTERN = re.compile(
+    r"(?P<prefix>[0-9A-Za-z\u4e00-\u9fff])\\\.(?=(?:\s|$|[)\]\u3002\uff0c\uff1b\uff1a\uff01\uff1f]))"
+)
+
 
 def _is_already_escaped(text: str) -> bool:
     """检测文本是否已经包含 MarkdownV2 转义字符。
@@ -534,68 +539,46 @@ def _unescape_markdown_v2(text: str) -> str:
     return re.sub(r"\\([_*\[\]()~`>#+=|{}.!:-])", r"\1", text)
 
 
-def _unescape_if_already_escaped(text: str) -> str:
-    """智能检测并清理预转义文本，特别保护代码块。
-
-    改进的分段处理策略：
-    1. 先识别已转义的代码块（\`\`\`...\`\`\` 和 \`...\`）
-    2. 对这些代码块先反转义边界反引号，变成正常代码块
-    3. 然后用正常的 CODE_SEGMENT_RE 识别代码块
-    4. 只对非代码块的普通文本进行反转义
-    5. 代码块内容保持转义状态（因为是代码本身）
-    6. 重新组合所有段落
-
-    Args:
-        text: 待处理的文本
-
-    Returns:
-        处理后的文本（如未检测到预转义，返回原文本）
-    """
+def _force_unescape_markdown(text: str) -> str:
+    """强制移除 MarkdownV2 转义，同时保护代码块语法不被破坏。"""
     if not text:
         return text
 
-    # 快速检测：如果没有任何转义字符，直接返回
-    if not _is_already_escaped(text):
-        return text
-
-    # 第一步：处理已转义的代码块，将边界反引号反转义
-    # 这样后续可以用正常的 CODE_SEGMENT_RE 识别它们
     processed = text
-
-    # 先标记所有已转义的代码块，用占位符替换
     code_blocks: list[str] = []
 
     def _preserve_code_block(match: re.Match[str]) -> str:
-        """保存代码块并返回占位符"""
+        """临时替换代码块，防止内部字符被错误反转义。"""
         block = match.group(0)
-        # 代码块边界的反引号需要反转义，但内容保持不变
-        # 例如：\`\`\`bash\npython -m vibego\_cli\`\`\`
-        #    -> ```bash\npython -m vibego\_cli```
         if block.startswith(r"\`\`\`"):
-            # 多行代码块
-            unescaped_block = block.replace(r"\`", "`", 6)  # 只替换前后各3个反引号
+            # 多行代码块保留内容，只修复边界反引号
+            unescaped_block = block.replace(r"\`", "`", 6)
         else:
-            # 单行代码块
-            unescaped_block = block.replace(r"\`", "`", 2)  # 只替换前后各1个反引号
+            # 单行代码块同理处理首尾反引号
+            unescaped_block = block.replace(r"\`", "`", 2)
 
         placeholder = f"__CODE_BLOCK_{len(code_blocks)}__"
         code_blocks.append(unescaped_block)
         return placeholder
 
     processed = _ESCAPED_CODE_BLOCK_PATTERN.sub(_preserve_code_block, processed)
+    processed = _unescape_markdown_v2(processed)
 
-    # 第二步：对非代码块部分进行反转义
-    if _is_already_escaped(processed):
-        processed = _unescape_markdown_v2(processed)
-
-    # 第三步：恢复代码块
-    for i, block in enumerate(code_blocks):
-        placeholder = f"__CODE_BLOCK_{i}__"
-        processed = processed.replace(placeholder, block)
+    for index, block in enumerate(code_blocks):
+        processed = processed.replace(f"__CODE_BLOCK_{index}__", block)
 
     return processed
 
 
+def _unescape_if_already_escaped(text: str) -> str:
+    """智能检测并清理预转义文本，必要时触发强制反转义。"""
+    if not text:
+        return text
+    if not _is_already_escaped(text):
+        return text
+    return _force_unescape_markdown(text)
+
+
 def _prepare_model_payload(text: str) -> str:
     if _IS_MARKDOWN_V2:
         cleaned = _unescape_if_already_escaped(text)
@@ -605,6 +588,47 @@ def _prepare_model_payload(text: str) -> str:
     return text
 
 
+def _optimize_markdown_v2_payload(payload: str) -> str:
+    """在保留 MarkdownV2 合法性的前提下，移除常见的冗余反斜杠。"""
+
+    if not payload or "\\" not in payload:
+        return payload
+
+    preserved_segments: list[str] = []
+
+    def _preserve(segment: re.Match[str]) -> str:
+        placeholder = f"__OPT_CODE_BLOCK_{len(preserved_segments)}__"
+        preserved_segments.append(segment.group(0))
+        return placeholder
+
+    stripped = CODE_SEGMENT_RE.sub(_preserve, payload)
+    stripped = _OPTIMIZE_CHECKBOX_PATTERN.sub("[ ]", stripped)
+
+    def _dot_replacer(match: re.Match[str]) -> str:
+        prefix = match.group("prefix")
+        return f"{prefix}."
+
+    stripped = _OPTIMIZE_DOT_PATTERN.sub(_dot_replacer, stripped)
+
+    for index, segment in enumerate(preserved_segments):
+        stripped = stripped.replace(f"__OPT_CODE_BLOCK_{index}__", segment)
+
+    return stripped
+
+
+def _prepare_model_payload_variants(text: str) -> tuple[str, Optional[str]]:
+    """返回首选与备用的 MarkdownV2 内容，首选尽量减少转义。"""
+
+    strict_payload = _prepare_model_payload(text)
+    if not _IS_MARKDOWN_V2:
+        return strict_payload, None
+
+    optimized_payload = _optimize_markdown_v2_payload(strict_payload)
+    if optimized_payload != strict_payload:
+        return optimized_payload, strict_payload
+    return strict_payload, None
+
+
 def _extract_bad_request_message(exc: TelegramBadRequest) -> str:
     message = getattr(exc, "message", None)
     if not message:
@@ -658,6 +682,7 @@ async def _send_with_markdown_guard(
     sender: Callable[[str], Awaitable[None]],
     *,
     raw_sender: Optional[Callable[[str], Awaitable[None]]] = None,
+    fallback_payload: Optional[str] = None,
 ) -> str:
     try:
         await sender(text)
@@ -666,6 +691,19 @@ async def _send_with_markdown_guard(
         if not _is_markdown_parse_error(exc):
             raise
 
+        if fallback_payload and fallback_payload != text:
+            try:
+                await sender(fallback_payload)
+                worker_log.debug(
+                    "Markdown 优化回退为严格转义版本",
+                    extra={"length": len(fallback_payload)},
+                )
+                return fallback_payload
+            except TelegramBadRequest as fallback_exc:
+                if not _is_markdown_parse_error(fallback_exc):
+                    raise
+                exc = fallback_exc
+
         sanitized: Optional[str]
         if _IS_MARKDOWN_V2:
             sanitized = _escape_markdown_v2(text)
@@ -695,7 +733,7 @@ async def _send_with_markdown_guard(
             raise
 
         # Markdown 彻底失败时退回纯文本发送，需要先清理 MarkdownV2 转义符号
-        fallback_payload = _unescape_if_already_escaped(text)
+        fallback_payload = _force_unescape_markdown(text)
         if fallback_payload != text:
             worker_log.debug(
                 "Markdown 降级为纯文本发送，已移除转义符号",
@@ -762,7 +800,11 @@ async def reply_large_text(
     """
     bot = current_bot()
     parse_mode_value = parse_mode if parse_mode is not None else _parse_mode_value()
-    prepared = text if preformatted else _prepare_model_payload(text)
+    if preformatted:
+        prepared = text
+        fallback_payload = None
+    else:
+        prepared, fallback_payload = _prepare_model_payload_variants(text)
 
     async def _send_formatted_message(payload: str) -> None:
         await bot.send_message(
@@ -779,6 +821,7 @@ async def reply_large_text(
             prepared,
             _send_formatted_message,
             raw_sender=_send_raw_message,
+            fallback_payload=fallback_payload,
         )
 
         worker_log.info(
@@ -797,10 +840,12 @@ async def reply_large_text(
         f"内容较长，已生成附件 `{attachment_name}`，请下载查看全文。"
     )
 
+    summary_prepared, summary_fallback = _prepare_model_payload_variants(summary_text)
     delivered_summary = await _send_with_markdown_guard(
-        summary_text,
+        summary_prepared,
         _send_formatted_message,
         raw_sender=_send_raw_message,
+        fallback_payload=summary_fallback,
     )
 
     document = BufferedInputFile(text.encode("utf-8"), filename=attachment_name)
@@ -1514,7 +1559,7 @@ async def _broadcast_worker_keyboard(bot: Bot) -> None:
             continue
 
         parse_mode = _parse_mode_value()
-        prepared = _prepare_model_payload(text)
+        prepared, fallback_payload = _prepare_model_payload_variants(text)
 
         async def _send_formatted(payload: str) -> None:
             await bot.send_message(
@@ -1537,6 +1582,7 @@ async def _broadcast_worker_keyboard(bot: Bot) -> None:
                 prepared,
                 _send_formatted,
                 raw_sender=_send_raw,
+                fallback_payload=fallback_payload,
             )
         except TelegramForbiddenError as exc:
             worker_log.warning("推送任务列表被拒绝：%s", exc, extra={**_session_extra(), "chat": chat_id})
@@ -2794,13 +2840,32 @@ async def _answer_with_markdown(
     *,
     reply_markup: InlineKeyboardMarkup | ReplyKeyboardMarkup | None = None,
 ) -> Optional[Message]:
-    prepared = _prepare_model_payload(text)
-    try:
-        sent = await message.answer(
-            prepared,
+    prepared, fallback_payload = _prepare_model_payload_variants(text)
+    sent_message: Optional[Message] = None
+
+    async def _send(payload: str) -> None:
+        nonlocal sent_message
+        sent_message = await message.answer(
+            payload,
             parse_mode=_parse_mode_value(),
             reply_markup=reply_markup,
         )
+
+    async def _send_raw(payload: str) -> None:
+        nonlocal sent_message
+        sent_message = await message.answer(
+            payload,
+            parse_mode=None,
+            reply_markup=reply_markup,
+        )
+
+    try:
+        await _send_with_markdown_guard(
+            prepared,
+            _send,
+            raw_sender=_send_raw,
+            fallback_payload=fallback_payload,
+        )
     except TelegramBadRequest as exc:
         worker_log.warning(
             "发送消息失败：%s",
@@ -2808,7 +2873,7 @@ async def _answer_with_markdown(
             extra={"chat": getattr(message.chat, "id", None)},
         )
         return None
-    return sent
+    return sent_message
 
 
 async def _edit_message_with_markdown(
@@ -2817,11 +2882,27 @@ async def _edit_message_with_markdown(
     *,
     reply_markup: InlineKeyboardMarkup | None = None,
 ) -> None:
-    prepared = _prepare_model_payload(text)
-    await callback.message.edit_text(
+    prepared, fallback_payload = _prepare_model_payload_variants(text)
+
+    async def _send(payload: str) -> None:
+        await callback.message.edit_text(
+            payload,
+            parse_mode=_parse_mode_value(),
+            reply_markup=reply_markup,
+        )
+
+    async def _send_raw(payload: str) -> None:
+        await callback.message.edit_text(
+            payload,
+            parse_mode=None,
+            reply_markup=reply_markup,
+        )
+
+    await _send_with_markdown_guard(
         prepared,
-        parse_mode=_parse_mode_value(),
-        reply_markup=reply_markup,
+        _send,
+        raw_sender=_send_raw,
+        fallback_payload=fallback_payload,
     )
 
 
@@ -2833,13 +2914,29 @@ async def _try_edit_message(
 ) -> bool:
     if message is None:
         return False
-    prepared = _prepare_model_payload(text)
-    try:
+    prepared, fallback_payload = _prepare_model_payload_variants(text)
+
+    async def _send(payload: str) -> None:
         await message.edit_text(
-            prepared,
+            payload,
             parse_mode=_parse_mode_value(),
             reply_markup=reply_markup,
         )
+
+    async def _send_raw(payload: str) -> None:
+        await message.edit_text(
+            payload,
+            parse_mode=None,
+            reply_markup=reply_markup,
+        )
+
+    try:
+        await _send_with_markdown_guard(
+            prepared,
+            _send,
+            raw_sender=_send_raw,
+            fallback_payload=fallback_payload,
+        )
         return True
     except TelegramBadRequest as exc:
         worker_log.info(

{vibego-0.2.35.dist-info → vibego-0.2.37.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: vibego
-Version: 0.2.35
+Version: 0.2.37
 Summary: vibego CLI：用于初始化与管理 Telegram Master Bot 的工具
 Author: Hypha
 License-Expression: LicenseRef-Proprietary

{vibego-0.2.35.dist-info → vibego-0.2.37.dist-info}/RECORD

@@ -1,4 +1,4 @@
-bot.py,sha256=bp1j-Dyp6FLsRkwbnxnFixFyJefFM4noJcpH_WIjzHQ,271364
+bot.py,sha256=RHE0IJcgiuXThExBQs4ZqxYF-Tfqpg3tIC5yYTGLr4o,274363
 logging_setup.py,sha256=gvxHi8mUwK3IhXJrsGNTDo-DR6ngkyav1X-tvlBF_IE,4613
 master.py,sha256=ZW4A3Gh0MUKFnfZX-VJ7OCNnBzlEcOWkPKYH92OfyKA,112967
 project_repository.py,sha256=UcthtSGOJK0cTE5bQCneo3xkomRG-kyc1N1QVqxeHIs,17577
@@ -427,14 +427,14 @@ tasks/fsm.py,sha256=rKXXLEieQQU4r2z_CZUvn1_70FXiZXBBugF40gpe_tQ,1476
 tasks/models.py,sha256=N_qqRBo9xMSV0vbn4k6bLBXT8C_dp_oTFUxvdx16ZQM,2459
 tasks/service.py,sha256=w_S_aWiVqRXzXEpimLDsuCCCX2lB5uDkff9aKThBw9c,41916
 telegram_markdown/__init__.py,sha256=bG3H9fWn5GfTqC6xvd49xbVdYWfSFeaX2nefweOYcWY,9757
-vibego_cli/__init__.py,sha256=_5XgkMeh6CuhPHHa6ulRX3B_Ja_dtOgJp9oowplquDg,311
+vibego_cli/__init__.py,sha256=n3G3v9lJLz5-SZnbrTB8h6FL6u1LF37J0JJ1pCcB_CM,311
 vibego_cli/__main__.py,sha256=qqTrYmRRLe4361fMzbI3-CqpZ7AhTofIHmfp4ykrrBY,158
 vibego_cli/config.py,sha256=VxkPJMq01tA3h3cOkH-z_tiP7pMgfSGGicRvUnCWkhI,3054
 vibego_cli/deps.py,sha256=1nRXI7Dd-S1hYE8DligzK5fIluQWETRUj4_OKL0DikQ,1419
 vibego_cli/main.py,sha256=X__NXwZnIDIFbdKSTbNyZgZHKcPlN0DQz9sqTI1aQ9E,12158
 vibego_cli/data/worker_requirements.txt,sha256=QSt30DSSSHtfucTFPpc7twk9kLS5rVLNTcvDiagxrZg,62
-vibego-0.2.35.dist-info/METADATA,sha256=EuDdzAX4rdAXErzufZdXP8aW326sLBDDj4oAPj7NvjM,10519
-vibego-0.2.35.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-vibego-0.2.35.dist-info/entry_points.txt,sha256=Lsy_zm-dlyxt8-9DL9blBReIwU2k22c8-kifr46ND1M,48
-vibego-0.2.35.dist-info/top_level.txt,sha256=rWDj9KERtbJL6Lar9Xa0O6dthaFSY_jc1WNpQgUrXCM,87
-vibego-0.2.35.dist-info/RECORD,,
+vibego-0.2.37.dist-info/METADATA,sha256=U_HSCRWJDsLvWIP2xu86nBcqrDhkJsOESZBJzLDnvSw,10519
+vibego-0.2.37.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+vibego-0.2.37.dist-info/entry_points.txt,sha256=Lsy_zm-dlyxt8-9DL9blBReIwU2k22c8-kifr46ND1M,48
+vibego-0.2.37.dist-info/top_level.txt,sha256=rWDj9KERtbJL6Lar9Xa0O6dthaFSY_jc1WNpQgUrXCM,87
+vibego-0.2.37.dist-info/RECORD,,

vibego_cli/__init__.py

@@ -7,6 +7,6 @@ from __future__ import annotations
 
 __all__ = ["main", "__version__"]
 
-__version__ = "0.2.35"
+__version__ = "0.2.37"
 
 from .main import main  # noqa: E402