loggier 0.1.1

Modern Python uygulamaları için log yönetim ve hata izleme kütüphanesi

Loggier

Loggier, Python tabanlı backend uygulamaları için geliştirilen, düşük maliyetli bir log yönetim sistemidir. Bu kütüphane, hataları ve logları yakalayıp işleyerek merkezi bir sunucuya göndermeyi sağlar.

Kurulum

pip install loggier

Hızlı Başlangıç

from loggier import Loggier

# Loggier'ı başlat
logger = Loggier(api_key="your_api_key_here")

# Temel log kayıtları
logger.info("Bu bir bilgi mesajıdır")
logger.warning("Bu bir uyarı mesajıdır")
logger.error("Bu bir hata mesajıdır")

# Bağlam ile log
logger.info("Kullanıcı oturum açtı", context={"user_id": 123, "username": "testuser"})

# Hata takibi
try:
    1 / 0
except Exception as e:
    logger.exception("Bir hata oluştu", exception=e)

# Context manager kullanımı
with logger.context(transaction="ödeme işlemi", user_id=123):
    # Bu blok içindeki tüm loglara bağlam bilgisi eklenir
    logger.info("Ödeme işlemi başlatıldı")
    logger.info("Ödeme işlemi tamamlandı")

Özellikler

• Esnek Log Seviyeleri: info, warning, error, critical
• Asenkron Gönderim: Uygulamanızı yavaşlatmadan log gönderimi
• Otomatik Bağlam Toplama: Çalışma zamanı bilgileri ve ortam verileri
• Hata İzleme: Otomatik istisna ve izleme yakalama
• Özel Etiketler ve Metadata: Loglarınızı özelleştirme
• Otomatik Yeniden Deneme: Bağlantı sorunlarında güvenilir log gönderimi
• Web Framework Entegrasyonları: Django, Flask, FastAPI için hazır entegrasyonlar

Gelişmiş Kullanım

Web Framework Entegrasyonları

Django Integration Guide for Loggier

This guide explains how to integrate Loggier with your Django project for request/response logging and error tracking.

Installation

First, ensure you have installed the Loggier package:

pip install loggier

Integration Options

Option 1: Direct Middleware (Recommended)

The simplest way to integrate Loggier with your Django project is to add LoggierDjangoMiddleware directly to your MIDDLEWARE setting in settings.py:

# settings.py
MIDDLEWARE = [
    'corsheaders.middleware.CorsMiddleware',  # If you use corsheaders
    'django.middleware.security.SecurityMiddleware',
    'django.contrib.sessions.middleware.SessionMiddleware',
    # ... other middleware
    'loggier.integrations.django.LoggierDjangoMiddleware',  # Add Loggier middleware
    # ... other middleware
]

# Configure Loggier
LOGGIER = {
    'API_KEY': 'your-api-key-here',
    'API_URL': 'http://localhost:8000/',  # Or your Loggier server URL
    'ENVIRONMENT': 'development',  # or 'production', 'staging', etc.
    'TAGS': ['django', 'web'],
    'CONTEXT': {
        'application': 'your-app-name'
    },
    'CAPTURE_REQUEST_DATA': True,
    'LOG_SLOW_REQUESTS': True,
    'SLOW_REQUEST_THRESHOLD': 1.0  # in seconds
}

Option 2: Custom Middleware Module (Legacy Approach)

If you need more control over the middleware initialization, you can create a custom middleware module:

# myapp/middleware.py
from loggier.integrations.django import LoggierDjango

# Create instance
loggier_django = LoggierDjango(
    api_url='http://localhost:8000/',
    api_key='your-api-key-here',
    environment='development',
    tags=['django', 'web'],
    # Other parameters...
)

# Get the middleware class
LoggierMiddleware = loggier_django.get_middleware()

Then in your settings.py:

MIDDLEWARE = [
    # ... other middleware
    'myapp.middleware.LoggierMiddleware',
    # ... other middleware
]

Configuration Options

The LOGGIER settings dictionary supports the following options:

Option	Description	Default
API_KEY	Your Loggier API key	Required
API_URL	Loggier API endpoint	https://api.loggier.com/api/ingest
ENVIRONMENT	Environment name (production, staging, etc.)	development
ASYNC_MODE	Whether to send logs asynchronously	True
MAX_BATCH_SIZE	Maximum batch size for async mode	100
FLUSH_INTERVAL	Flush interval in seconds for async mode	5
TAGS	List of tags to add to all logs	[]
CONTEXT	Dictionary of context data to add to all logs	{}
LOG_LEVEL	Minimum log level to send	info
CAPTURE_REQUEST_DATA	Whether to capture request data	True
LOG_SLOW_REQUESTS	Whether to log slow requests	True
SLOW_REQUEST_THRESHOLD	Threshold for slow requests in seconds	1.0

What Gets Logged

With the default configuration, Loggier will automatically log:

1. Request Start: Basic information about each incoming request
2. Request End: Response status, timing, and size
3. Exceptions: Detailed exception information with traceback
4. Slow Requests: Warning logs for requests that exceed the threshold

Manual Logging

You can also use the Loggier client directly in your views:

from loggier import Loggier

logger = Loggier(
    api_key='your-api-key',
    api_url='http://localhost:8000/',
    environment='development'
)

def my_view(request):
    # Log information
    logger.info("Processing user data", context={"user_id": request.user.id})

    try:
        # Your code here
        result = process_data()
        logger.info("Data processed successfully", context={"result_size": len(result)})
    except Exception as e:
        # Log exceptions
        logger.exception("Error processing data", exception=e)
        raise

    return HttpResponse("Success")

Troubleshooting

Parameter Errors

If you encounter errors like TypeError: __init__() got an unexpected keyword argument 'context', it means your installed version of Loggier doesn't match the expected API. Check your installed version with:

pip show loggier

And either:

1. Update to the latest version: pip install --upgrade loggier
2. Modify your middleware to match the API of your installed version

Middleware Import Errors

Make sure the middleware path in your MIDDLEWARE setting exactly matches where you've defined the middleware class. If using the legacy approach, ensure you're referencing LoggierMiddleware (not LoggierDjango):

# CORRECT:
'myapp.middleware.LoggierMiddleware',  # Notice it's LoggierMiddleware, not LoggierDjango

# INCORRECT:
'myapp.middleware.LoggierDjango',  # This won't work

Performance Considerations

• Loggier uses asynchronous logging by default, which minimizes performance impact
• For high-traffic applications, consider adjusting the MAX_BATCH_SIZE and FLUSH_INTERVAL
• In production environments, make sure to call flush() during application shutdown to ensure all logs are sent