error-handling-patterns（错误处理模式）

Error Handling Patterns

Use this skill when implementing error handling for features or API endpoints.

Checklist

Exception Handling Rules

• Never use bare except - always specify exception type
• Catch specific exceptions first, then broader ones
• Use domain-specific exception classes with meaningful names
• Log errors before re-raising or returning error responses

Domain-Specific Exceptions

• Create custom exception classes that extend built-in exceptions
• Use meaningful exception names that describe the error condition
• Include helpful error messages with context

class ModelNotFoundError(ValueError):
    """Raised when a requested model is not found."""
    pass

class InsufficientMemoryError(RuntimeError):
    """Raised when GPU memory is insufficient for operation."""
    pass

API Response Patterns

• Always use Pydantic schemas for responses (never raw dicts)
• Include error details in response schema
• Use appropriate HTTP status codes

HTTP Status Codes

Use the correct status code for each situation:

• 200 - Success (operation completed successfully)
• 400 - Bad Request (invalid input, validation failure)
• 404 - Not Found (resource doesn't exist)
• 409 - Conflict (resource already exists, state conflict)
• 500 - Internal Server Error (unexpected failure)

Example Pattern

from app.services import logger_service
from app.schemas.responses import ErrorResponse, SuccessResponse

logger = logger_service.get_logger(__name__, category='API')

async def generate_image(config: GenerateConfig, db: Session):
    try:
        result = await service.generate_image(config, db)
        return SuccessResponse(data=result)
    except ValueError as error:
        logger.error(f"Generation failed: {error}")
        raise HTTPException(status_code=400, detail=str(error))
    except torch.cuda.OutOfMemoryError:
        logger.error("Out of GPU memory")
        raise HTTPException(status_code=500, detail="Insufficient GPU memory")
    except Exception as error:
        logger.exception(f"Unexpected error: {error}")
        raise HTTPException(status_code=500, detail="Internal server error")

Validation

• Verify all exception types are specific (no bare except:)
• Verify all API responses use Pydantic schemas
• Verify HTTP status codes match the error conditions
• Verify errors are logged before raising/returning