Logging

Introduction

Use the stdlib logging module — never print() in production. Levels: DEBUG, INFO, WARNING, ERROR, CRITICAL.

Understanding the topic

Core concepts to understand:

• Get a logger per module: logging.getLogger(__name__).
• Configure once at app entry (basicConfig or dictConfig).
• Structured logs (JSON) for production.
• Use logger.exception() in except blocks.

Syntax reference

Visual flow / code:

import logging

logging.basicConfig(
    level=logging.INFO,
    format="%(asctime)s %(levelname)s %(name)s %(message)s",
)
logger = logging.getLogger(__name__)

try:
    risky_op()
except Exception:
    logger.exception("risky_op failed")  # logs traceback

logger.info("user %s logged in", user_id)

Execution workflow

Real-world use

Production stacks ship JSON logs to ELK/Loki/Datadog. Libraries like structlog and loguru are popular alternatives.

Best practices

• Never print() in libraries/services.
• Log at INFO for business events, DEBUG for diagnostics.
• Use structured logging in prod.

Common mistakes

• String-formatting in log calls: use %s, not f-strings, to avoid evaluating when level filters out.

Hands-on exercise

Interview preparation — practice these questions:

• Logger vs print?
• What's structured logging?
• logger.exception vs error?

Summary

In summary: Logging > print, always. Structured = searchable + parseable.