aiecs 1.2.2__py3-none-any.whl → 1.3.3__py3-none-any.whl

aiecs/tools/apisource/reliability/error_handler.py

@@ -0,0 +1,362 @@
+"""
+Smart Error Handler for API Providers
+
+Provides intelligent error handling with:
+- Classification of retryable vs non-retryable errors
+- Exponential backoff retry strategy
+- Agent-friendly error messages with recovery suggestions
+- Detailed error context and history
+"""
+
+import logging
+import time
+from datetime import datetime
+from typing import Any, Callable, Dict, List, Optional, Tuple
+
+logger = logging.getLogger(__name__)
+
+
+class SmartErrorHandler:
+    """
+    Intelligent error handler with retry logic and recovery suggestions.
+    
+    Automatically classifies errors, applies appropriate retry strategies,
+    and provides actionable recovery suggestions for AI agents.
+    """
+    
+    # Error types that can be retried
+    RETRYABLE_ERRORS = [
+        'timeout',
+        'connection',
+        'rate limit',
+        'rate_limit',
+        '429',  # Too Many Requests
+        '500',  # Internal Server Error
+        '502',  # Bad Gateway
+        '503',  # Service Unavailable
+        '504',  # Gateway Timeout
+        'temporary',
+        'transient'
+    ]
+    
+    # Error types that should not be retried
+    NON_RETRYABLE_ERRORS = [
+        'authentication',
+        'auth',
+        'api key',
+        'unauthorized',
+        '401',  # Unauthorized
+        '403',  # Forbidden
+        'invalid',
+        'not found',
+        '404',  # Not Found
+        'bad request',
+        '400'   # Bad Request
+    ]
+    
+    def __init__(
+        self,
+        max_retries: int = 3,
+        backoff_factor: float = 2.0,
+        initial_delay: float = 1.0,
+        max_delay: float = 30.0
+    ):
+        """
+        Initialize error handler.
+        
+        Args:
+            max_retries: Maximum number of retry attempts
+            backoff_factor: Multiplier for exponential backoff
+            initial_delay: Initial delay in seconds before first retry
+            max_delay: Maximum delay between retries in seconds
+        """
+        self.max_retries = max_retries
+        self.backoff_factor = backoff_factor
+        self.initial_delay = initial_delay
+        self.max_delay = max_delay
+    
+    def execute_with_retry(
+        self,
+        operation_func: Callable[[], Any],
+        operation_name: str = "operation",
+        provider_name: str = "provider"
+    ) -> Dict[str, Any]:
+        """
+        Execute an operation with intelligent retry logic.
+        
+        Args:
+            operation_func: Function to execute (should take no arguments)
+            operation_name: Name of the operation for logging
+            provider_name: Name of the provider for logging
+        
+        Returns:
+            Dictionary with:
+                - success: bool
+                - data: result data if successful
+                - error: error details if failed
+                - retry_info: retry attempt information
+        """
+        last_error = None
+        retry_info = {
+            'attempts': 0,
+            'errors': [],
+            'recovery_suggestions': []
+        }
+        
+        for attempt in range(self.max_retries):
+            retry_info['attempts'] = attempt + 1
+            
+            try:
+                # Execute the operation
+                result = operation_func()
+                
+                # Success!
+                if attempt > 0:
+                    logger.info(
+                        f"{provider_name}.{operation_name} succeeded after "
+                        f"{attempt + 1} attempts"
+                    )
+                
+                return {
+                    'success': True,
+                    'data': result,
+                    'retry_info': retry_info
+                }
+            
+            except Exception as e:
+                last_error = e
+                error_msg = str(e).lower()
+                error_type = self._classify_error(e, error_msg)
+                
+                # Record this error
+                retry_info['errors'].append({
+                    'attempt': attempt + 1,
+                    'error': str(e),
+                    'error_type': error_type,
+                    'timestamp': datetime.utcnow().isoformat()
+                })
+                
+                # Determine if we should retry
+                is_retryable = self._is_retryable(error_msg, error_type)
+                is_last_attempt = (attempt == self.max_retries - 1)
+                
+                if not is_retryable:
+                    logger.warning(
+                        f"{provider_name}.{operation_name} failed with non-retryable "
+                        f"error: {error_type}"
+                    )
+                    break
+                
+                if is_last_attempt:
+                    logger.error(
+                        f"{provider_name}.{operation_name} failed after "
+                        f"{self.max_retries} attempts"
+                    )
+                    break
+                
+                # Calculate wait time with exponential backoff
+                wait_time = min(
+                    self.initial_delay * (self.backoff_factor ** attempt),
+                    self.max_delay
+                )
+                
+                logger.info(
+                    f"{provider_name}.{operation_name} attempt {attempt + 1} failed "
+                    f"({error_type}), retrying in {wait_time:.1f}s..."
+                )
+                
+                time.sleep(wait_time)
+        
+        # All retries exhausted or non-retryable error
+        retry_info['recovery_suggestions'] = self._generate_recovery_suggestions(
+            last_error,
+            operation_name,
+            provider_name
+        )
+        
+        return {
+            'success': False,
+            'error': {
+                'type': type(last_error).__name__,
+                'message': str(last_error),
+                'details': self._parse_error_details(last_error),
+                'is_retryable': self._is_retryable(str(last_error).lower(), 
+                                                   self._classify_error(last_error, str(last_error).lower()))
+            },
+            'retry_info': retry_info
+        }
+    
+    def _classify_error(self, error: Exception, error_msg: str) -> str:
+        """
+        Classify error type.
+        
+        Args:
+            error: The exception object
+            error_msg: Error message in lowercase
+        
+        Returns:
+            Error classification string
+        """
+        # Check exception type
+        error_class = type(error).__name__.lower()
+        
+        if 'timeout' in error_class or 'timeout' in error_msg:
+            return 'timeout'
+        elif 'connection' in error_class or 'connection' in error_msg:
+            return 'connection'
+        elif 'rate' in error_msg and 'limit' in error_msg:
+            return 'rate_limit'
+        elif '429' in error_msg:
+            return 'rate_limit'
+        elif 'auth' in error_msg or '401' in error_msg or '403' in error_msg:
+            return 'authentication'
+        elif 'not found' in error_msg or '404' in error_msg:
+            return 'not_found'
+        elif 'bad request' in error_msg or '400' in error_msg:
+            return 'invalid_parameter'
+        elif '500' in error_msg or '502' in error_msg or '503' in error_msg or '504' in error_msg:
+            return 'server_error'
+        elif 'invalid' in error_msg:
+            return 'invalid_parameter'
+        else:
+            return 'unknown'
+    
+    def _is_retryable(self, error_msg: str, error_type: str) -> bool:
+        """
+        Determine if an error is retryable.
+        
+        Args:
+            error_msg: Error message in lowercase
+            error_type: Classified error type
+        
+        Returns:
+            True if error is retryable
+        """
+        # Check non-retryable first (higher priority)
+        for non_retryable in self.NON_RETRYABLE_ERRORS:
+            if non_retryable in error_msg or non_retryable in error_type:
+                return False
+        
+        # Check retryable
+        for retryable in self.RETRYABLE_ERRORS:
+            if retryable in error_msg or retryable in error_type:
+                return True
+        
+        # Default: retry server errors and unknown errors
+        return error_type in ['server_error', 'timeout', 'connection', 'rate_limit']
+    
+    def _parse_error_details(self, error: Exception) -> Dict[str, Any]:
+        """
+        Extract detailed information from error.
+        
+        Args:
+            error: The exception object
+        
+        Returns:
+            Dictionary with error details
+        """
+        details = {
+            'class': type(error).__name__,
+            'message': str(error)
+        }
+        
+        # Extract HTTP status code if available
+        if hasattr(error, 'response') and hasattr(error.response, 'status_code'):
+            details['status_code'] = error.response.status_code
+            if hasattr(error.response, 'text'):
+                details['response_body'] = error.response.text[:500]  # Limit size
+        
+        # Extract additional context
+        if hasattr(error, '__cause__') and error.__cause__:
+            details['cause'] = str(error.__cause__)
+        
+        return details
+    
+    def _generate_recovery_suggestions(
+        self,
+        error: Exception,
+        operation_name: str,
+        provider_name: str
+    ) -> List[str]:
+        """
+        Generate actionable recovery suggestions for the error.
+        
+        Args:
+            error: The exception object
+            operation_name: Name of the failed operation
+            provider_name: Name of the provider
+        
+        Returns:
+            List of recovery suggestion strings
+        """
+        suggestions = []
+        error_msg = str(error).lower()
+        error_type = self._classify_error(error, error_msg)
+        
+        if error_type == 'authentication':
+            suggestions.extend([
+                f"Check that your {provider_name.upper()} API key is valid and properly configured",
+                f"Verify the API key environment variable {provider_name.upper()}_API_KEY is set",
+                "Confirm the API key has not expired or been revoked",
+                f"Ensure you have the necessary permissions for the {operation_name} operation"
+            ])
+        
+        elif error_type == 'rate_limit':
+            suggestions.extend([
+                "Wait before making more requests (rate limit exceeded)",
+                f"Consider using a different provider if available",
+                "Reduce the frequency of requests or implement request batching",
+                "Check if you can upgrade your API plan for higher rate limits",
+                "Enable caching to reduce API calls"
+            ])
+        
+        elif error_type == 'not_found':
+            suggestions.extend([
+                f"Verify the resource ID or parameter values are correct for {provider_name}",
+                f"Use the search or list operations to find valid resource IDs",
+                "Check that the resource exists and is accessible with your credentials",
+                f"Review the {provider_name} API documentation for correct resource identifiers"
+            ])
+        
+        elif error_type == 'invalid_parameter':
+            suggestions.extend([
+                f"Check the operation schema for {operation_name} to see required parameters",
+                "Review parameter types and value formats",
+                f"Use get_operation_schema('{operation_name}') to see valid parameters and examples",
+                "Verify parameter names are spelled correctly"
+            ])
+        
+        elif error_type == 'timeout':
+            suggestions.extend([
+                "Try again with a smaller date range or result limit",
+                "Increase the timeout setting in the provider configuration",
+                "Check your network connection",
+                "The API service may be experiencing high load, try again later"
+            ])
+        
+        elif error_type == 'connection':
+            suggestions.extend([
+                "Check your internet connection",
+                "Verify the API endpoint is accessible",
+                "Check if there are any firewall or proxy issues",
+                "The API service may be temporarily unavailable"
+            ])
+        
+        elif error_type == 'server_error':
+            suggestions.extend([
+                "The API service is experiencing issues, try again later",
+                "Check the API status page for known outages",
+                "Try a different provider if available",
+                "Reduce the complexity of your request"
+            ])
+        
+        else:
+            suggestions.extend([
+                f"Review the error message for specific details",
+                f"Check the {provider_name} API documentation for {operation_name}",
+                "Verify all required parameters are provided",
+                "Try a simpler query to isolate the issue"
+            ])
+        
+        return suggestions
+