feat(webchat): refactor message parsing logic and integrate new parsing function

2026-02-25 17:14:02 +08:00
parent 53ae8cd7cf
commit f8f7e6d57a
2 changed files with 157 additions and 72 deletions
@@ -4,6 +4,7 @@ import shutil
 import uuid
 from collections.abc import Awaitable, Callable, Sequence
 from pathlib import Path
+from typing import Any

 from astrbot.core.db.po import Attachment
 from astrbot.core.message.components import (
@@ -19,6 +20,10 @@ from astrbot.core.message.message_event_result import MessageChain

 AttachmentGetter = Callable[[str], Awaitable[Attachment | None]]
 AttachmentInserter = Callable[[str, str, str], Awaitable[Attachment | None]]
+ReplyHistoryGetter = Callable[
+    [Any],
+    Awaitable[tuple[list[dict], str | None, str | None] | None],
+]

 MEDIA_PART_TYPES = {"image", "record", "file", "video"}

@@ -35,6 +40,127 @@ def webchat_message_parts_have_content(message_parts: list[dict]) -> bool:
    )


+async def parse_webchat_message_parts(
+    message_parts: list,
+    *,
+    strict: bool = False,
+    include_empty_plain: bool = False,
+    verify_media_path_exists: bool = True,
+    reply_history_getter: ReplyHistoryGetter | None = None,
+    current_depth: int = 0,
+    max_reply_depth: int = 0,
+    cast_reply_id_to_str: bool = True,
+) -> tuple[list, list[str], bool]:
+    """Parse webchat message parts into components/text parts.
+
+    Returns:
+        tuple[list, list[str], bool]:
+            (components, plain_text_parts, has_non_reply_content)
+    """
+    components = []
+    text_parts: list[str] = []
+    has_content = False
+
+    for part in message_parts:
+        if not isinstance(part, dict):
+            if strict:
+                raise ValueError("message part must be an object")
+            continue
+
+        part_type = str(part.get("type", "")).strip()
+        if part_type == "plain":
+            text = str(part.get("text", ""))
+            if text or include_empty_plain:
+                components.append(Plain(text=text))
+                text_parts.append(text)
+            if text:
+                has_content = True
+            continue
+
+        if part_type == "reply":
+            message_id = part.get("message_id")
+            if message_id is None:
+                if strict:
+                    raise ValueError("reply part missing message_id")
+                continue
+
+            reply_chain = []
+            reply_message_str = str(part.get("selected_text", ""))
+            sender_id = None
+            sender_name = None
+
+            if reply_message_str:
+                reply_chain = [Plain(text=reply_message_str)]
+            elif (
+                reply_history_getter
+                and current_depth < max_reply_depth
+                and message_id is not None
+            ):
+                reply_info = await reply_history_getter(message_id)
+                if reply_info:
+                    reply_parts, sender_id, sender_name = reply_info
+                    (
+                        reply_chain,
+                        reply_text_parts,
+                        _,
+                    ) = await parse_webchat_message_parts(
+                        reply_parts,
+                        strict=strict,
+                        include_empty_plain=include_empty_plain,
+                        verify_media_path_exists=verify_media_path_exists,
+                        reply_history_getter=reply_history_getter,
+                        current_depth=current_depth + 1,
+                        max_reply_depth=max_reply_depth,
+                        cast_reply_id_to_str=cast_reply_id_to_str,
+                    )
+                    reply_message_str = "".join(reply_text_parts)
+
+            reply_id = str(message_id) if cast_reply_id_to_str else message_id
+            components.append(
+                Reply(
+                    id=reply_id,
+                    message_str=reply_message_str,
+                    chain=reply_chain,
+                    sender_id=sender_id,
+                    sender_nickname=sender_name,
+                )
+            )
+            continue
+
+        if part_type not in MEDIA_PART_TYPES:
+            if strict:
+                raise ValueError(f"unsupported message part type: {part_type}")
+            continue
+
+        path = part.get("path")
+        if not path:
+            if strict:
+                raise ValueError(f"{part_type} part missing path")
+            continue
+
+        file_path = Path(str(path))
+        if verify_media_path_exists and not file_path.exists():
+            if strict:
+                raise ValueError(f"file not found: {file_path!s}")
+            continue
+
+        file_path_str = (
+            str(file_path.resolve()) if verify_media_path_exists else str(file_path)
+        )
+        has_content = True
+        if part_type == "image":
+            components.append(Image.fromFileSystem(file_path_str))
+        elif part_type == "record":
+            components.append(Record.fromFileSystem(file_path_str))
+        elif part_type == "video":
+            components.append(Video.fromFileSystem(file_path_str))
+        else:
+            filename = str(part.get("filename", "")).strip() or file_path.name
+            components.append(File(name=filename, file=file_path_str))
+
+    return components, text_parts, has_content
+
+
 async def build_webchat_message_parts(
    message_payload: str | list,
    *,
@@ -192,7 +318,13 @@ async def build_message_chain_from_payload(
        get_attachment_by_id=get_attachment_by_id,
        strict=strict,
    )
-    return webchat_message_parts_to_message_chain(message_parts, strict=strict)
+    components, _, has_content = await parse_webchat_message_parts(
+        message_parts,
+        strict=strict,
+    )
+    if strict and (not components or not has_content):
+        raise ValueError("Message content is empty (reply only is not allowed)")
+    return MessageChain(chain=components)


 async def create_attachment_part_from_existing_file(
@@ -9,14 +9,6 @@ from typing import Any
 from astrbot import logger
 from astrbot.core import db_helper
 from astrbot.core.db.po import PlatformMessageHistory
-from astrbot.core.message.components import (
-    File,
-    Image,
-    Plain,
-    Record,
-    Reply,
-    Video,
-)
 from astrbot.core.message.message_event_result import MessageChain
 from astrbot.core.platform import (
    AstrBotMessage,
@@ -29,7 +21,10 @@ from astrbot.core.platform.astr_message_event import MessageSesion
 from astrbot.core.utils.astrbot_path import get_astrbot_data_path

 from ...register import register_platform_adapter
-from .message_parts_helper import message_chain_to_storage_message_parts
+from .message_parts_helper import (
+    message_chain_to_storage_message_parts,
+    parse_webchat_message_parts,
+)
 from .webchat_event import WebChatMessageEvent
 from .webchat_queue_mgr import WebChatQueueMgr, webchat_queue_mgr

@@ -175,72 +170,30 @@ class WebChatAdapter(Platform):
        Returns:
            tuple[list, list[str]]: (消息组件列表, 纯文本列表)
        """
-        components = []
-        text_parts = []

-        for part in message_parts:
-            part_type = part.get("type")
-            if part_type == "plain":
-                text = part.get("text", "")
-                components.append(Plain(text=text))
-                text_parts.append(text)
-            elif part_type == "reply":
-                message_id = part.get("message_id")
-                reply_chain = []
-                reply_message_str = part.get("selected_text", "")
-                sender_id = None
-                sender_name = None
+        async def get_reply_parts(
+            message_id: Any,
+        ) -> tuple[list[dict], str | None, str | None] | None:
+            history = await self._get_message_history(message_id)
+            if not history or not history.content:
+                return None

-                if reply_message_str:
-                    reply_chain = [Plain(text=reply_message_str)]
+            reply_parts = history.content.get("message", [])
+            if not isinstance(reply_parts, list):
+                return None

-                # recursively get the content of the referenced message, if selected_text is empty
-                if not reply_message_str and depth < max_depth and message_id:
-                    history = await self._get_message_history(message_id)
-                    if history and history.content:
-                        reply_parts = history.content.get("message", [])
-                        if isinstance(reply_parts, list):
-                            (
-                                reply_chain,
-                                reply_text_parts,
-                            ) = await self._parse_message_parts(
-                                reply_parts,
-                                depth=depth + 1,
-                                max_depth=max_depth,
-                            )
-                            reply_message_str = "".join(reply_text_parts)
-                        sender_id = history.sender_id
-                        sender_name = history.sender_name
-
-                components.append(
-                    Reply(
-                        id=message_id,
-                        chain=reply_chain,
-                        message_str=reply_message_str,
-                        sender_id=sender_id,
-                        sender_nickname=sender_name,
-                    )
-                )
-            elif part_type == "image":
-                path = part.get("path")
-                if path:
-                    components.append(Image.fromFileSystem(path))
-            elif part_type == "record":
-                path = part.get("path")
-                if path:
-                    components.append(Record.fromFileSystem(path))
-            elif part_type == "file":
-                path = part.get("path")
-                if path:
-                    filename = part.get("filename") or (
-                        os.path.basename(path) if path else "file"
-                    )
-                    components.append(File(name=filename, file=path))
-            elif part_type == "video":
-                path = part.get("path")
-                if path:
-                    components.append(Video.fromFileSystem(path))
+            return reply_parts, history.sender_id, history.sender_name

+        components, text_parts, _ = await parse_webchat_message_parts(
+            message_parts,
+            strict=False,
+            include_empty_plain=True,
+            verify_media_path_exists=False,
+            reply_history_getter=get_reply_parts,
+            current_depth=depth,
+            max_reply_depth=max_depth,
+            cast_reply_id_to_str=False,
+        )
        return components, text_parts

    async def convert_message(self, data: tuple) -> AstrBotMessage: