feat(backend): 文件处理和语义分割 API 更新

- chunks API: 支持语义分割模式和 embedding 配置
- files API: 文件异步处理优化

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

backend/app/api/v1/chunks/__init__.py

@@ -17,7 +17,7 @@ from app.core.crud import CRUDBase
 from app.core.logging import log_success, log_failure
 from app.models.models import Chunk, File
 from app.schemas.chunk import ChunkResponse
-from app.schemas.chunk import ChunkCreateSchema
+from app.schemas.chunk import ChunkCreateSchema, ChunkUpdateSchema
 from app.services.text_splitter.splitter import get_splitter
 from markitdown import MarkItDown
 
@@ -209,7 +209,7 @@ async def list_chunks(
         limit=page_size,
         filters=filters,
         order_by="created_at",
-        descending=True
+        descending=False
     )
 
     chunk_responses = [ChunkResponse.model_validate(c) for c in chunks]
@@ -239,7 +239,7 @@ async def get_chunk(
 async def update_chunk(
     project_id: UUID,
     chunk_id: UUID,
-    chunk: ChunkCreateSchema,
+    chunk: ChunkUpdateSchema,
     db: AsyncSession = Depends(get_db)
 ):
     """Update chunk"""

backend/app/api/v1/files/__init__.py

@@ -158,7 +158,7 @@ async def upload_file(
                         # txt, md 等直接读取
                         text_content = file_path_obj.read_text(encoding='utf-8')
 
-                    # 保存到 ready 目录
+                    # 保存到 ready 目录，使用 {uuid}.md 格式
                     ready_dir = get_project_ready_dir(str(project_id_val))
                     ready_filename = f"{file_id}.md"
                     ready_path = ready_dir / ready_filename