inspari_logging 0.1.8

Inspari logging utilities

This repository provides Python logging utilities.

Configuration

The inspari.logging module provides unified interface for loading logging configuration files in app code and from gunicorn. In addition, a range of utilities (service name prefixing, command line colors etc.) are provided.

An example logging configuration file is bundled (example_logging_config.json), along with an example of the usage in an application context (example_usage.py).

Stream logs

The streamlogs command line utility provides a simple way to stream logs in realtime. To enable log collection, add a handler as part of your logging configuration,

    "handlers": {
        "web": {
            "class": "inspari.logging.AzureBlobStorageHandler",
            "formatter": "simple",
            "load_dot_env": "true",
            "env_key": "ABS_CONNECTION_STRING",
            "log_local": "true"
        },
        ...
    },
    ...
    "root": {
        "level": "INFO",
        "handlers": [
            "web",
            ...
        ]
    },

and configure the logs as demonstrated in example_usage.py. The storage account to use for streaming must be specified via environment variables. In the example case, a connection string is used. Hence, for local development purposes, it is recommended to create a .env file with the content like,

ABS_CONNECTION_STRING=REDACTED

Note that by default, logs are written to container called logs, so you must create a container with this name, if it doesn't already exist. If you now run,

poetry run steamlogs

you should be getting logs in (near) realtime from all services connected to the account.

Development

Create a new Python environment with all dependencies installed,

poetry sync

That's it! You can validate that the environment is setup correctly by running the tests,

poetry run coverage run -m pytest

Deployment

Update the version in pyproject.toml, run poetry lock, and add a new entry in CHANGELOG.md.

Automatic (preferred)

Merge the changes into master (using PRs). Create a new tag. If the new version is 1.1.1, the new tag would be v1.1.1.

The tag creation will trigger automated deployment of the package to PyPi, as well as synchronization of the code with the public GitHub mirror repository.

Manual (not recommended)

Build the project via Poetry,

poetry build

and push it to pypi,

poetry publish -u __token__ -p $INSPARI_PYPI_TOKEN