feat(pkg-py): Add new message_content() and message_content_chunk() (#96)

* feat(contents_shinychat): Add new contents_shinychat() and contents_shinychat_chunk()

These new functions are the singledispatch equivalent of the previous normalize_message() and normalize_message_chunk(). Although not quite as flexible as the previous strategies pattern, it's much simpler (implementation wise and also for those registering new methods) and aligns much better with the vision for the R package

* Apply suggestions from code review

Co-authored-by: Copilot <[email protected]>

* Fix format check

* Rename

* Add ChatMessage to types module. Add types and playwright module to top-level namespace

* Fix for Anthropic; cleanup

* Just require python 3.11 with Anthropic

* Update pkg-py/src/shinychat/_chat_normalize.py

Co-authored-by: Copilot <[email protected]>

* Just require python 3.11 with Anthropic

* Just require python 3.11 with Anthropic

* Add a test for registering custom objects

* Ignore missing anthropic imports when type checking

* fml

* rename (again)

* rename (again)

---------

Co-authored-by: Copilot <[email protected]>

Author: cpsievert

pkg-py/CHANGELOG.md

@@ -7,10 +7,14 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [UNRELEASED]
 
-* The chat input no longer submits incomplete text when the user has activated IME completions (e.g. while typing in Japanese or Chinese). (#85)
+### New features
+
+* Added new `message_content()` and `message_content_chunk()` generic (`singledispatch`) functions. These functions aren't intended to be called directly by users, but instead, provide an opportunity to teach `Chat.append_message()`/`Chat.append_message_stream()` to extract message contents from different types of objects. (#96)
 
+### Bug fixes
+
+* The chat input no longer submits incomplete text when the user has activated IME completions (e.g. while typing in Japanese or Chinese). (#85)
 
 ## [0.1.0] - 2025-08-07
 
 This first release of the `shinychat` package simply copies the `Chat` and `MarkdownStream` components exactly as they are in version 1.4.0 of `shiny`. Future versions of `shiny` will import these components from `shinychat`. By maintaining these components via a separate library, we can ship features more quickly and independently of `shiny`.
-

pkg-py/src/shinychat/__init__.py

@@ -1,4 +1,15 @@
+from . import playwright, types
 from ._chat import Chat, chat_ui
+from ._chat_normalize import message_content, message_content_chunk
 from ._markdown_stream import MarkdownStream, output_markdown_stream
 
-__all__ = ["Chat", "chat_ui", "MarkdownStream", "output_markdown_stream"]
+__all__ = [
+    "Chat",
+    "chat_ui",
+    "MarkdownStream",
+    "output_markdown_stream",
+    "message_content",
+    "message_content_chunk",
+    "types",
+    "playwright",
+]

pkg-py/src/shinychat/_chat.py

@@ -52,9 +52,9 @@
     is_chatlas_chat_client,
     set_chatlas_state,
 )
-from ._chat_normalize import normalize_message, normalize_message_chunk
+from ._chat_normalize import message_content, message_content_chunk
 from ._chat_provider_types import (
-    AnthropicMessage,
+    AnthropicMessage,  # pyright: ignore[reportAttributeAccessIssue]
     GoogleMessage,
     LangChainMessage,
     OllamaMessage,
@@ -636,7 +636,7 @@ async def append_message(
             self._pending_messages.append((message, False, "append", None))
             return
 
-        msg = normalize_message(message)
+        msg = message_content(message)
         msg = await self._transform_message(msg)
         if msg is None:
             return
@@ -753,7 +753,7 @@ async def _append_message_chunk(
         self._current_stream_id = stream_id
 
         # Normalize various message types into a ChatMessage()
-        msg = normalize_message_chunk(message)
+        msg = message_content_chunk(message)
 
         if operation == "replace":
             self._current_stream_message = (