gcp-log

Python library to format logs as GCP-compatible JSON.

Features:

• Works with standard logging.
• Supports extra for structured logging.
• 100% compatible with the official spec.
• Non-invasive, doesn't start any threads or processes, doesn't make network or system calls.
• Zero configuration.
• Fast.
• Pure Python.
• Type-safe.
• Zero dependency.
• Optional integration with orjson for better performance.

Installation

python3 -m pip install gcp-log

Optional: If you also install orjson, gcp-log will automatically use it instead of stdlib json for faster serialization:

python3 -m pip install orjson

Usage

To use it, simply set gcp_log.Formatter as the formatter for the logger:

import logging
import gcp_log

logger = logging.getLogger()
handler = logging.StreamHandler()
formatter = gcp_log.Formatter()
handler.setFormatter(formatter)
logger.addHandler(handler)

And then you can use the logger:

logger.info('aragorn', extra=dict(father='arathorn'))

The output:

{"message":"aragorn","severity":"INFO","timestamp":"2022-03-17T10:09:58.393124+00:00Z","logging.googleapis.com/sourceLocation":{"file":"/full/path/to/example.py.py","line":24,"function":"test_gcp_formatter"},"father":"arathorn"}

A human-readable version:

{
    "message": "aragorn",
    "severity": "INFO",
    "timestamp": "2022-03-17T10:09:58.393124+00:00Z",
    "logging.googleapis.com/sourceLocation": {
        "file": "/full/path/to/example.py.py",
        "line": 24,
        "function": "test_gcp_formatter",
    },
    "father": "arathorn",
}