Logging

Logging

This Wiki page is used to describe our best practices for logging scripts.

General Principles

• Logging should be included in every script.
• Logging should capture errors handled by the script and and output messages.

Best Practices

• Log files should be named like the following: <name_of_script>_YYYY_MM_DD.log
• Log files should be written to the logs directory, as set up by a user in their local repository.
• The logs directory and log files should be excluded from any commits to the repository. Instead, log files from scripts run against production should be stored in OneDrive - Smithsonian Institution\Documents - OCIO-Digital Platforms-CAAS\Data Cleanup Projects\ArchivesSpace.
• Log files should clearly indicate where errors were raised using the logging library's ERROR status when available.
• Log files can include any additional info the programmer thinks is useful for debugging, general info (such as DEBUG, INFO, WARNING)

Logging Libraries

• loguru (preferred)

Code example (using loguru)

import loguru

logger.remove()
log_path = Path('../logs', 'delete_dometadata_{time:YYYY-MM-DD}.log')
logger.add(str(log_path), format="{time}-{level}: {message}")

Example log file

2024-11-21T16:35:26.253085-0500-INFO: Updated /repositories/32/digital_objects/222717
2024-11-21T16:35:26.316023-0500-INFO: Updated /repositories/32/digital_objects/222718
2024-11-21T16:35:26.360491-0500-ERROR: get_object() - Unable to retrieve object with provided URI: {'error': 'DigitalObject not found'}