Metadata-Version: 2.3
Name: macwinnie-enhanced-logging
Version: 0.1.3
Summary: Batteries-included logging for Python applications.
Author: macwinnie
Author-email: macwinnie <dev@macwinnie.me>
Requires-Dist: structlog>=25.5.0
Requires-Python: >=3.12
Project-URL: Homepage, https://github.com/macwinnie/py-enhanced-logging
Project-URL: Issues, https://github.com/macwinnie/py-enhanced-logging/issues
Description-Content-Type: text/markdown

# macwinnie-enhanced-logging

Last built version: `v0.1.3`

Batteries-included logging for Python applications.

`macwinnie-enhanced-logging` provides a clean, consistent logging setup built on top of:

- Python’s built-in `logging` module (so third-party libraries keep working normally)
- [`structlog`](https://www.structlog.org/) (so *your* logs are structured, predictable, and easy to analyze)

It supports both **developer-friendly console logs** and **structured JSON logs** suitable for production environments like Kubernetes.

Further info: see [GitHub Repo](https://github.com/macwinnie/py-enhanced-logging) and [Online-Documentation](https://macwinnie.github.io/py-enhanced-logging).

---

## Features

- Works with **standard Python logging**
- Structured logging with **structlog**
- **Human-friendly console output** for development
- **JSON output** for log pipelines and observability stacks
- Optional **callsite information** (file/function/line)
- Built-in **context variables** (request_id, job_id, etc.)
- Optional **speaking logs** for critical events
- Zero boilerplate for most applications

---

## Installation

```bash
pip install macwinnie-enhanced-logging
```

Requires Python **3.12+**.

---

## Quick Start

### 1. Configure logging once in your application entry point

```python
from enhanced_logging import configure_logging, get_logger

configure_logging()

log = get_logger(__name__)
log.info("startup", _fmt="Starting {app}…", app="demo")
```

### 2. Use the logger anywhere in your code

```python
from enhanced_logging import get_logger

log = get_logger(__name__)

log.info("something_happened", user_id=123)
```

---

## Output Formats

Two output formats are supported.

### Human format (default)

Developer-friendly console output:

```
2025-01-01T10:00:00Z [info] Starting api on 127.0.0.1:8080 component=api host=127.0.0.1 port=8080
```

### JSON format

Machine-friendly structured output:

```json
{
  "timestamp": "2025-01-01T10:00:00Z",
  "level": "info",
  "event": "startup",
  "msg": "Starting api on 127.0.0.1:8080",
  "component": "api",
  "host": "127.0.0.1",
  "port": 8080
}
```

Switch formats using an environment variable:

```bash
LOG_FORMAT=json
```

---

## `_fmt`: Human-Readable Messages with Structured Logs

You can provide `_fmt` to generate a human-readable message while still keeping structured fields.

Example:

```python
log.info(
    "startup",
    _fmt="Starting {component} on {host}:{port}",
    component="api",
    host="127.0.0.1",
    port=8080,
)
```

Behavior:

- **human format** → replaces the event message
- **json format** → keeps `event` and adds `msg`

This allows you to combine **structured logging** with **readable messages**.

---

## Context Variables

You can attach fields to **all logs within a scope** using `log_context`.

Example:

```python
from enhanced_logging import get_logger, log_context

log = get_logger(__name__)

with log_context(request_id="req-123", user_id=42):
    log.info("handling_request")
    do_work()
    log.info("done")
```

All logs inside the block will automatically include:

```
request_id=req-123 user_id=42
```

This is ideal for:

- request IDs
- job IDs
- tenant identifiers
- correlation IDs

---

## Environment Variables

Configuration is primarily done through environment variables.

| Variable | Description | Default |
|--------|--------|--------|
| `LOG_FORMAT` | `human` or `json` | `human` |
| `LOG_LEVEL` | Logging level | `INFO` |
| `LOG_UTC` | Use UTC timestamps | `1` |
| `LOG_KEEP_HANDLERS` | Keep existing root handlers | `0` |
| `LOG_CALLSITE` | `auto`, `1`, or `0` | `auto` |
| `LOG_APP_NAME` | Optional application name | unset |
| `LOG_SPEAK` | Enable spoken logs | `0` |
| `LOG_SPEAK_LEVEL` | Minimum level to speak | `ERROR` |
| `LOG_SPEAK_CMD` | Custom speech command | auto |

---

## Callsite Information

Callsite fields include:

- filename
- function name
- line number

Behavior:

- Automatically enabled when `LOG_LEVEL=DEBUG`
- Can be forced via:

```bash
LOG_CALLSITE=1
```

Disable explicitly:

```bash
LOG_CALLSITE=0
```

---

## Speaking Logs (Optional)

`macwinnie-enhanced-logging` can optionally **speak important log messages aloud**.

This can be useful for:

- CI pipelines
- long-running batch jobs
- alerting during development

Enable speaking:

```bash
LOG_SPEAK=1
```

Only errors and above are spoken by default.

Change threshold:

```bash
LOG_SPEAK_LEVEL=WARNING
```

Platform support:

| Platform | Engine |
|--------|--------|
| macOS | `say` |
| Linux | `espeak` |
| Windows | PowerShell SAPI |

If no speech engine is available, speaking is silently disabled.

You can also call the helper directly:

```python
from enhanced_logging import speak

speak("Deployment finished successfully")
```

---

## Configuration Example

```python
from enhanced_logging import configure_logging, LogConfig

configure_logging(
    LogConfig(
        format="json",
        level="INFO",
        app_name="my-service"
    )
)
```

---

## Integration with Third-Party Libraries

Because `macwinnie-enhanced-logging` configures the **root logger**, libraries that use:

```python
logging.getLogger(...)
```

will automatically emit logs through the same handler.

This means:

- one logging configuration for your entire application
- consistent formatting across libraries

---

## Best Practices

- Call `configure_logging()` **once at application startup**
- Use `get_logger(__name__)` in each module
- Prefer structured fields over string concatenation
- Use `_fmt` when you want readable console output

Example:

```python
log.info(
    "user_login",
    _fmt="User {user_id} logged in",
    user_id=123
)
```

---

## Acknowledgements

Author of this package is [@macwinnie](https://chaos.social/@macwinnie).

This project builds on:

- Python `logging` module
- [`structlog`](https://www.structlog.org/)

which provide the foundation for structured logging in Python.
