Python Error Handling

Python Error Handling

Build robust Python applications with proper input validation, meaningful exceptions, and graceful failure handling. Good error handling makes debugging easier and systems more reliable.

When to Use This Skill

• Validating user input and API parameters
• Designing exception hierarchies for applications
• Handling partial failures in batch operations
• Converting external data to domain types
• Building user-friendly error messages
• Implementing fail-fast validation patterns

Core Concepts

1. Fail Fast

Validate inputs early, before expensive operations. Report all validation errors at once when possible.

2. Meaningful Exceptions

Use appropriate exception types with context. Messages should explain what failed, why, and how to fix it.

3. Partial Failures

In batch operations, don't let one failure abort everything. Track successes and failures separately.

4. Preserve Context

Chain exceptions to maintain the full error trail for debugging.

5. Scope with resilience and resource-management skills

• Fail-fast vs fail-safe: python-resilience uses defaults when an operation is optional or degradable (e.g. empty recommendations). Keep fail-fast validation and raise for money, security, invariants, and authoritative writes. Do not decorate the same code path with both instincts unless the product explicitly defines which failures may degrade.

• Broad except: In transactions and context managers, except Exception: that rolls back or cleans up and then raise is intentional�any failure must undo the unit of work. See python-resource-management. That is not the anti-pattern of except Exception: pass or vague catches without re-raise.

Quick Start

def fetch_page(url: str, page_size: int) -> Page:
    if not url:
        raise ValueError("'url' is required")
    if not 1 <= page_size <= 100:
        raise ValueError(f"'page_size' must be 1-100, got {page_size}")
    # Now safe to proceed...

Fundamental Patterns

Pattern 1: Early Input Validation

Validate all inputs at API boundaries before any processing begins.

def process_order(
    order_id: str,
    quantity: int,
    discount_percent: float,
) -> OrderResult:
    """Process an order with validation."""
    # Validate required fields
    if not order_id:
        raise ValueError("'order_id' is required")

    # Validate ranges
    if quantity <= 0:
        raise ValueError(f"'quantity' must be positive, got {quantity}")

    if not 0 <= discount_percent <= 100:
        raise ValueError(
            f"'discount_percent' must be 0-100, got {discount_percent}"
        )

    # Validation passed, proceed with processing
    return _process_validated_order(order_id, quantity, discount_percent)

Pattern 2: Convert to Domain Types Early

Parse strings and external data into typed domain objects at system boundaries.

from enum import Enum

class OutputFormat(Enum):
    JSON = "json"
    CSV = "csv"
    PARQUET = "parquet"

def parse_output_format(value: str) -> OutputFormat:
    """Parse string to OutputFormat enum.

    Args:
        value: Format string from user input.

    Returns:
        Validated OutputFormat enum member.

    Raises:
        ValueError: If format is not recognized.
    """
    try:
        return OutputFormat(value.lower())
    except ValueError:
        valid_formats = [f.value for f in OutputFormat]
        raise ValueError(
            f"Invalid format '{value}'. "
            f"Valid options: {', '.join(valid_formats)}"
        )

# Usage at API boundary
def export_data(data: list[dict], format_str: str) -> bytes:
    output_format = parse_output_format(format_str)  # Fail fast
    # Rest of function uses typed OutputFormat
    ...

Pattern 3: Pydantic for Complex Validation

Use Pydantic models for structured input validation with automatic error messages.

Required fields do not need Field(...) (ellipsis). A type without a default is already required; put constraints on Field without using ... as the default. This matches the fastapi skill and keeps models consistent in FastAPI apps.

from pydantic import BaseModel, Field, ValidationError, field_validator

class CreateUserInput(BaseModel):
    """Input model for user creation."""

    email: str = Field(min_length=5, max_length=255)
    name: str = Field(min_length=1, max_length=100)
    age: int = Field(ge=0, le=150)

    @field_validator("email")
    @classmethod
    def validate_email_format(cls, v: str) -> str:
        if "@" not in v or "." not in v.split("@")[-1]:
            raise ValueError("Invalid email format")
        return v.lower()

    @field_validator("name")
    @classmethod
    def normalize_name(cls, v: str) -> str:
        return v.strip().title()

# Usage
try:
    user_input = CreateUserInput(
        email="user@example.com",
        name="john doe",
        age=25,
    )
except ValidationError as e:
    # Pydantic provides detailed error information
    print(e.errors())

Pattern 4: Map Errors to Standard Exceptions

Use Python's built-in exception types appropriately, adding context as needed.

# Good: Specific exception with context
raise ValueError(f"'page_size' must be 1-100, got {page_size}")

# Avoid: Generic exception, no context
raise Exception("Invalid parameter")

Advanced Patterns

Pattern 5: Custom Exceptions with Context

Create domain-specific exceptions that carry structured information.

class ApiError(Exception):
    """Base exception for API errors."""

    def __init__(
        self,
        message: str,
        status_code: int,
        response_body: str | None = None,
    ) -> None:
        self.status_code = status_code
        self.response_body = response_body
        super().__init__(message)

class RateLimitError(ApiError):
    """Raised when rate limit is exceeded."""

    def __init__(self, retry_after: int) -> None:
        self.retry_after = retry_after
        super().__init__(
            f"Rate limit exceeded. Retry after {retry_after}s",
            status_code=429,
        )

# Usage
def handle_response(response: Response) -> dict:
    match response.status_code:
        case 200:
            return response.json()
        case 401:
            raise ApiError("Invalid credentials", 401)
        case 404:
            raise ApiError(f"Resource not found: {response.url}", 404)
        case 429:
            retry_after = int(response.headers.get("Retry-After", 60))
            raise RateLimitError(retry_after)
        case code if 400 <= code < 500:
            raise ApiError(f"Client error: {response.text}", code)
        case code if code >= 500:
            raise ApiError(f"Server error: {response.text}", code)

Pattern 6: Exception Chaining

Preserve the original exception when re-raising to maintain the debug trail.

import httpx

class ServiceError(Exception):
    """High-level service operation failed."""
    pass

def upload_file(path: str) -> str:
    """Upload file and return URL."""
    try:
        with open(path, "rb") as f:
            response = httpx.post("https://upload.example.com", files={"file": f})
            response.raise_for_status()
            return response.json()["url"]
    except FileNotFoundError as e:
        raise ServiceError(f"Upload failed: file not found at '{path}'") from e
    except httpx.HTTPStatusError as e:
        raise ServiceError(
            f"Upload failed: server returned {e.response.status_code}"
        ) from e
    except httpx.RequestError as e:
        raise ServiceError(f"Upload failed: network error") from e

Pattern 7: Batch Processing with Partial Failures

Never let one bad item abort an entire batch. Track results per item.

from dataclasses import dataclass

@dataclass
class BatchResult[T]:
    """Results from batch processing."""

    succeeded: dict[int, T]  # index -> result
    failed: dict[int, Exception]  # index -> error

    @property
    def success_count(self) -> int:
        return len(self.succeeded)

    @property
    def failure_count(self) -> int:
        return len(self.failed)

    @property
    def all_succeeded(self) -> bool:
        return len(self.failed) == 0

def process_batch(items: list[Item]) -> BatchResult[ProcessedItem]:
    """Process items, capturing individual failures.

    Args:
        items: Items to process.

    Returns:
        BatchResult with succeeded and failed items by index.
    """
    succeeded: dict[int, ProcessedItem] = {}
    failed: dict[int, Exception] = {}

    for idx, item in enumerate(items):
        try:
            result = process_single_item(item)
            succeeded[idx] = result
        except Exception as e:
            failed[idx] = e

    return BatchResult(succeeded=succeeded, failed=failed)

# Caller handles partial results
result = process_batch(items)
if not result.all_succeeded:
    logger.warning(
        f"Batch completed with {result.failure_count} failures",
        failed_indices=list(result.failed.keys()),
    )

Pattern 8: Progress Reporting for Long Operations

Provide visibility into batch progress without coupling business logic to UI.

from collections.abc import Callable

ProgressCallback = Callable[[int, int, str], None]  # current, total, status

def process_large_batch(
    items: list[Item],
    on_progress: ProgressCallback | None = None,
) -> BatchResult:
    """Process batch with optional progress reporting.

    Args:
        items: Items to process.
        on_progress: Optional callback receiving (current, total, status).
    """
    total = len(items)
    succeeded = {}
    failed = {}

    for idx, item in enumerate(items):
        if on_progress:
            on_progress(idx, total, f"Processing {item.id}")

        try:
            succeeded[idx] = process_single_item(item)
        except Exception as e:
            failed[idx] = e

    if on_progress:
        on_progress(total, total, "Complete")

    return BatchResult(succeeded=succeeded, failed=failed)

Best Practices Summary

1. Validate early - Check inputs before expensive operations
2. Use specific exceptions - Prefer ValueError / TypeError over vague handlers; avoid except Exception: pass. Transactional except Exception + cleanup + raise is intentional�see Scope with resilience and resource-management skills above.
3. Include context - Messages should explain what, why, and how to fix
4. Convert types at boundaries - Parse strings to enums/domain types early
5. Chain exceptions - Use raise ... from e to preserve debug info
6. Handle partial failures - Don't abort batches on single item errors
7. Use Pydantic - For complex input validation with structured errors
8. Document failure modes - Docstrings should list possible exceptions
9. Log with context - Include IDs, counts, and other debugging info
10. Test error paths - Verify exceptions are raised correctly