From 841b4df09def08fefa55a82475b649b603871029 Mon Sep 17 00:00:00 2001
From: Haoxuan Jiang <haoxuanjng@gmail.com>
Date: Mon, 19 Oct 2026 16:26:16 +0800
Subject: [PATCH] fix(lark): preserve incoming rich-text code blocks (#2626)

---
 .../adapters/lark/message_converter.py        |  2 +-
 src/langbot/pkg/platform/sources/lark.py      |  2 +-
 .../platform/test_lark_inbound_code_blocks.py | 78 +++++++++++++++++++
 3 files changed, 80 insertions(+), 2 deletions(-)
 create mode 100644 tests/unit_tests/platform/test_lark_inbound_code_blocks.py

diff --git a/src/langbot/pkg/platform/adapters/lark/message_converter.py b/src/langbot/pkg/platform/adapters/lark/message_converter.py
index 510246cf8..40db36bde 100644
--- a/src/langbot/pkg/platform/adapters/lark/message_converter.py
+++ b/src/langbot/pkg/platform/adapters/lark/message_converter.py
@@ -184,7 +184,7 @@ class LarkMessageConverter(abstract_platform_adapter.AbstractMessageConverter):
         normalized = LarkMessageConverter._normalize_inbound_content(message, message_content)
         for ele in normalized:
             tag = ele.get('tag')
-            if tag in {'text', 'md'}:
+            if tag in {'text', 'md', 'code_block'}:
                 text = ele.get('text') or ''
                 if text:
                     components.append(platform_message.Plain(text=text))
diff --git a/src/langbot/pkg/platform/sources/lark.py b/src/langbot/pkg/platform/sources/lark.py
index c4bd5c8a9..4dc600c8f 100644
--- a/src/langbot/pkg/platform/sources/lark.py
+++ b/src/langbot/pkg/platform/sources/lark.py
@@ -726,7 +726,7 @@ class LarkMessageConverter(abstract_platform_adapter.AbstractMessageConverter):
             ]
 
         for ele in message_content['content']:
-            if ele['tag'] == 'text':
+            if ele['tag'] in {'text', 'code_block'}:
                 lb_msg_list.append(platform_message.Plain(text=ele['text']))
             elif ele['tag'] == 'at':
                 lb_msg_list.append(platform_message.At(target=ele['user_name']))
diff --git a/tests/unit_tests/platform/test_lark_inbound_code_blocks.py b/tests/unit_tests/platform/test_lark_inbound_code_blocks.py
new file mode 100644
index 000000000..965abce56
--- /dev/null
+++ b/tests/unit_tests/platform/test_lark_inbound_code_blocks.py
@@ -0,0 +1,78 @@
+"""Lark rich-text code blocks must reach the bot as text instead of being dropped."""
+
+import json
+from types import SimpleNamespace
+from unittest.mock import MagicMock
+
+import pytest
+
+from langbot.pkg.platform.adapters.lark.message_converter import LarkMessageConverter
+from langbot.pkg.platform.sources.lark import LarkMessageConverter as LegacyLarkMessageConverter
+from langbot_plugin.api.entities.builtin.platform import message as platform_message
+
+
+@pytest.fixture(params=[LegacyLarkMessageConverter, LarkMessageConverter], ids=['legacy', 'omni'])
+def converter(request):
+    return request.param
+
+
+def post_message(content):
+    return SimpleNamespace(
+        message_id='post-msg',
+        message_type='post',
+        create_time='1714000000000',
+        content=json.dumps({'title': '', 'content': content}),
+        mentions=[],
+    )
+
+
+@pytest.mark.asyncio
+@pytest.mark.parametrize(
+    'code_block',
+    [
+        {'tag': 'code_block', 'language': 'GO', 'text': 'func main() int64 {\n    return 0\n}'},
+        {'tag': 'code_block', 'language': 'PYTHON', 'text': 'def hello():\n    print("你好")\n'},
+        {'tag': 'code_block', 'text': '    indented\n\nlast line'},
+    ],
+    ids=['official-go-example', 'python-unicode', 'no-language'],
+)
+async def test_inbound_post_preserves_code_block_text(converter, code_block):
+    api_client = MagicMock()
+
+    chain = await converter.target2yiri(post_message([[code_block]]), api_client)
+
+    assert len(chain) == 2
+    assert isinstance(chain[0], platform_message.Source)
+    assert chain[0].id == 'post-msg'
+    assert isinstance(chain[1], platform_message.Plain)
+    assert chain[1].text == code_block['text']
+    assert api_client.mock_calls == []
+
+
+@pytest.mark.asyncio
+async def test_inbound_post_preserves_code_between_text_and_mentions(converter):
+    api_client = MagicMock()
+    chain = await converter.target2yiri(
+        post_message(
+            [
+                [{'tag': 'at', 'user_id': 'ou_bot', 'user_name': 'Bot'}, {'tag': 'text', 'text': 'Explain: '}],
+                [{'tag': 'code_block', 'language': 'PYTHON', 'text': 'print(1 + 2)'}],
+                [{'tag': 'text', 'text': ' What is the result?'}],
+            ]
+        ),
+        api_client,
+    )
+
+    assert [type(component) for component in chain] == [
+        platform_message.Source,
+        platform_message.At,
+        platform_message.Plain,
+        platform_message.Plain,
+        platform_message.Plain,
+    ]
+    assert [component.text for component in chain if isinstance(component, platform_message.Plain)] == [
+        'Explain: ',
+        'print(1 + 2)',
+        ' What is the result?',
+    ]
+    assert api_client.mock_calls == []