loggier 0.1.3

Modern Python uygulamaları için log yönetim ve hata izleme kütüphanesi

PyLogHub/Loggier Documentation

Overview

PyLogHub (Loggier) is a comprehensive logging solution for Python applications, with specialized integrations for web frameworks like Django, Flask, and FastAPI. It offers a flexible and powerful way to collect, process, and analyze logs from your applications with features like asynchronous logging, local caching, performance monitoring, and web framework integrations.

Table of Contents

1. Installation
2. Core Concepts
3. Basic Usage
4. Advanced Configuration
5. Django Integration
6. Flask Integration
7. FastAPI Integration
8. Performance Monitoring
9. Error Tracking
10. Architecture
11. API Reference

Installation

Install using pip:

pip install loggier

Core Concepts

PyLogHub/Loggier is built around these core concepts:

• Client: The main Loggier class that handles log collection and sending
• Handlers: Components that process and deliver logs (API, Async, Cache)
• Formatters: Components that format logs for output
• Context: A system for adding and organizing contextual information
• Integrations: Framework-specific components for seamless integration

Basic Usage

Simple example of using Loggier:

from loggier import Loggier

# Create a Loggier instance
logger = Loggier(
    api_key="your-api-key",
    environment="development",
    service_name="my-service"
)

# Log basic messages
logger.info("Application started")
logger.warning("Something might be wrong", context={"user_id": 123})
logger.error("Something went wrong", context={"order_id": "ORD-123456"})

# Log exceptions
try:
    # Some code that might raise an exception
    result = 1 / 0
except Exception as e:
    logger.exception("Error calculating result", exception=e)

# Use context for adding structured information
with logger.context(user_id=123, transaction_id="TX12345"):
    logger.info("Processing transaction")
    try:
        # Transaction processing code
        logger.info("Transaction completed")
    except Exception as e:
        logger.exception("Transaction failed", exception=e)

# Make sure to flush logs before application exits
logger.flush()

Advanced Configuration

Loggier provides many configuration options:

from loggier import Loggier

logger = Loggier(
    api_key="your-api-key",
    project_name="my-project",
    environment="production",
    service_name="payment-service",
    api_url="https://api.loggier.com/api/ingest",  # Custom API endpoint
    async_mode=True,                              # Send logs asynchronously
    capture_uncaught=True,                        # Capture uncaught exceptions
    log_level=logging.INFO,                       # Minimum log level
    enable_caching=True,                          # Cache logs when offline
    cache_dir="/path/to/cache",                   # Custom cache directory
    max_batch_size=20,                            # Logs per batch
    flush_interval=5,                             # Auto-flush interval (seconds)
    sensitive_fields=["password", "credit_card"], # Fields to mask
    max_retries=3,                                # API retry attempts
    http_timeout=5,                               # API timeout (seconds)
    network_check_interval=60,                    # Network check interval
    tags=["api", "payments"],                     # Global tags
    enable_performance_monitoring=True            # Enable performance monitoring
)

Django Integration

Loggier offers a powerful Django integration that can track requests, responses, and errors.

Basic Setup

Add the middleware to your Django settings:

# settings.py
MIDDLEWARE = [
    # other middleware...
    'loggier.integrations.django.LoggierDjangoMiddleware',
    # other middleware...
]

# Configure Loggier
LOGGIER = {
    'API_KEY': 'your-api-key-here',
    'API_URL': 'https://api.loggier.com/api/ingest',
    'ENVIRONMENT': 'production',
    'TAGS': ['django', 'web'],
    'CAPTURE_REQUEST_DATA': True,
    'LOG_SLOW_REQUESTS': True,
    'SLOW_REQUEST_THRESHOLD': 1.0,  # seconds
    'STATUSES_TO_TRACK': [500, 400, 401, 403],  # Status codes to track
    'CAPTURE_REQUEST_BODY': False,  # Be careful with sensitive data
    'CAPTURE_RESPONSE_BODY': True,  # Capture response bodies for errors
    'MAX_BODY_SIZE': 16384,  # Max body size to capture (16KB)
    'INCLUDE_STACKTRACE': True  # Include stacktrace for errors
}

Tracking Specific Endpoints

Use the track_endpoint decorator to track specific views or endpoints:

from loggier.integrations.django import track_endpoint
from django.http import JsonResponse

# Track this endpoint with custom settings
@track_endpoint(
    tags=['critical', 'payment'],
    capture_request_body=True,
    capture_response_body=True,
    statuses=[200, 201, 400, 500]
)
def payment_process(request):
    # This endpoint will have complete request/response tracking
    return JsonResponse({"status": "success"})

Django Rest Framework Integration

With Django Rest Framework, you can track specific ViewSet methods:

from rest_framework import viewsets
from loggier.integrations.django import track_endpoint

class UserViewSet(viewsets.ModelViewSet):
    queryset = User.objects.all()
    serializer_class = UserSerializer

    # Track only the list action
    @track_endpoint(tags=['user', 'list'])
    def list(self, request, *args, **kwargs):
        return super().list(request, *args, **kwargs)

    # Track the create action with body capturing
    @track_endpoint(
        tags=['user', 'create'],
        capture_request_body=True,
        capture_response_body=True
    )
    def create(self, request, *args, **kwargs):
        return super().create(request, *args, **kwargs)

Complete Transaction Tracking

The Django middleware can track the complete context of HTTP transactions:

1. Request Information:

   • HTTP method, path, query parameters
   • Headers (filtered for sensitive data)
   • Request body (if enabled)
   • User information
   • View/controller information

2. Response Information:

   • Status code
   • Response headers
   • Response time
   • Response body (if enabled)

3. Error Information:

   • Exception type and message
   • Full traceback
   • Application state at time of error

4. Custom Context:

   • Add custom context using logger.context.update_global()

Example Log Output (Error):

{
  "level": "ERROR",
  "message": "Server error: GET /api/v1/templates/ - 500",
  "timestamp": "2025-03-09T11:41:29.000Z",
  "context": {
    "request_id": "3c4e6f8g-9h0i-1j2k-3l4m-5n6o7p8q9r0s",
    "transaction": {
      "request": {
        "method": "GET",
        "path": "/api/v1/templates/",
        "query_params": {
          "limit": "100",
          "offset": "0"
        },
        "user": {
          "user_id": 5,
          "username": "[e-mail]"
        }
      },
      "response": {
        "status_code": 500,
        "content_type": "text/plain; charset=utf-8",
        "traceback": "Traceback (most recent call last):...",
        "exception": {
          "type": "ValueError",
          "value": "Uncaught"
        }
      }
    }
  }
}

Flask Integration

Loggier provides integration with Flask:

from flask import Flask
from loggier.integrations.flask import LoggierFlask

app = Flask(__name__)

# Initialize Loggier
loggier = LoggierFlask(
    api_key="your-api-key",
    environment="production",
    capture_request_data=True,
    log_slow_requests=True,
    slow_request_threshold=1.0
)

# Register with Flask
loggier.init_app(app)

@app.route('/')
def index():
    return "Hello World!"

if __name__ == '__main__':
    app.run()

FastAPI Integration

Loggier integrates with FastAPI:

from fastapi import FastAPI
from loggier.integrations.fastapi import LoggierFastAPI

app = FastAPI()

# Initialize Loggier
loggier = LoggierFastAPI(
    api_key="your-api-key",
    environment="production",
    capture_request_data=True,
    log_slow_requests=True,
    slow_request_threshold=1.0
)

# Register with FastAPI
loggier.init_app(app)

@app.get("/")
def read_root():
    return {"Hello": "World"}

Performance Monitoring

Loggier includes tools for performance monitoring:

from loggier import Loggier

logger = Loggier(
    api_key="your-api-key",
    environment="production",
    enable_performance_monitoring=True
)

# Track function performance
@logger.trace_function(threshold_ms=100)
def process_data(data):
    # Function execution time will be tracked
    # If it exceeds 100ms, it will be logged
    return data

# Track HTTP requests
@logger.trace_http(threshold_ms=1000)
def fetch_external_api():
    import requests
    return requests.get("https://api.example.com/data")

# Track database operations
@logger.trace_database(threshold_ms=50)
def get_user(user_id):
    # Database operation time will be tracked
    return User.objects.get(id=user_id)

Error Tracking

Loggier provides comprehensive error tracking capabilities:

from loggier import Loggier

logger = Loggier(
    api_key="your-api-key",
    environment="production",
    capture_uncaught=True  # Automatically capture uncaught exceptions
)

# Manually log exceptions
try:
    # Some code that might raise an exception
    result = 1 / 0
except Exception as e:
    logger.exception(
        "Division error",
        exception=e,
        context={
            "operation": "division",
            "numerator": 1,
            "denominator": 0
        }
    )

# Capture exceptions in functions
@logger.capture_exceptions
def risky_function():
    # If this function raises an exception, it will be logged
    return 1 / 0

Architecture

PyLogHub/Loggier uses a modular architecture:

1. Client (Loggier): Main interface for applications
2. Handlers:
   • APIHandler: Communicates with the PyLogHub API
   • AsyncHandler: Handles asynchronous log sending
   • CacheHandler: Manages local caching of logs
3. Formatters:
   • JSONFormatter: Formats logs as JSON
4. Utils:
   • Context: Manages contextual information
   • Error: Handles exception capturing
   • Network: Monitors network connectivity

API Reference

Loggier Class

The main client class for interacting with PyLogHub.

Constructor

Loggier(
    api_key: str,
    project_name: Optional[str] = None,
    environment: str = "development",
    api_url: Optional[str] = None,
    service_name: Optional[str] = None,
    async_mode: bool = True,
    capture_uncaught: bool = True,
    log_level: int = logging.INFO,
    enable_caching: bool = True,
    cache_dir: Optional[str] = None,
    max_batch_size: int = 20,
    flush_interval: int = 5,
    sensitive_fields: Optional[List[str]] = None,
    max_retries: int = 3,
    http_timeout: int = 5,
    network_check_interval: int = 60,
    tags: Optional[List[str]] = None,
    enable_performance_monitoring: bool = True
)

Logging Methods

# Basic logging methods
log(level: str, message: str, exception: Optional[Exception] = None, context: Optional[Dict[str, Any]] = None, tags: Optional[List[str]] = None) -> bool
debug(message: str, exception: Optional[Exception] = None, context: Optional[Dict[str, Any]] = None, tags: Optional[List[str]] = None) -> bool
info(message: str, exception: Optional[Exception] = None, context: Optional[Dict[str, Any]] = None, tags: Optional[List[str]] = None) -> bool
warning(message: str, exception: Optional[Exception] = None, context: Optional[Dict[str, Any]] = None, tags: Optional[List[str]] = None) -> bool
error(message: str, exception: Optional[Exception] = None, context: Optional[Dict[str, Any]] = None, tags: Optional[List[str]] = None) -> bool
critical(message: str, exception: Optional[Exception] = None, context: Optional[Dict[str, Any]] = None, tags: Optional[List[str]] = None) -> bool
exception(message: str, exception: Optional[Exception] = None, context: Optional[Dict[str, Any]] = None, tags: Optional[List[str]] = None) -> bool

# Context management
context(**kwargs) -> Context

# Performance monitoring decorators
trace_function(name: Optional[str] = None, threshold_ms: int = 500, tags: Optional[List[str]] = None, include_args: bool = False, include_return: bool = False, log_level: str = "INFO") -> Callable
trace_http(name: Optional[str] = None, threshold_ms: int = 1000, tags: Optional[List[str]] = None, mask_headers: Optional[List[str]] = None, mask_params: Optional[List[str]] = None, include_body: bool = False, log_level: str = "INFO") -> Callable
trace_database(name: Optional[str] = None, threshold_ms: int = 100, tags: Optional[List[str]] = None, include_params: bool = False, log_level: str = "INFO") -> Callable

# Utility methods
flush(timeout: Optional[float] = None) -> bool
get_stats() -> Dict[str, Any]
shutdown(timeout: Optional[float] = None) -> None

Django Integration

LoggierDjangoMiddleware

# In settings.py
MIDDLEWARE = [
    'loggier.integrations.django.LoggierDjangoMiddleware',
    # other middleware...
]

LOGGIER = {
    'API_KEY': 'your-api-key',
    'API_URL': 'https://api.loggier.com/api/ingest',
    'ENVIRONMENT': 'production',
    'STATUSES_TO_TRACK': [500, 400, 401, 403],
    'CAPTURE_REQUEST_BODY': False,
    'CAPTURE_RESPONSE_BODY': True,
    'MAX_BODY_SIZE': 16384,
    'INCLUDE_STACKTRACE': True
}

track_endpoint Decorator

from loggier.integrations.django import track_endpoint

@track_endpoint(
    tags=None,                   # List of tags to add
    statuses=None,               # Status codes to track
    capture_request_body=False,  # Whether to capture request body
    capture_response_body=False  # Whether to capture response body
)
def my_view(request):
    # View implementation
    return HttpResponse("Hello")

Flask Integration

from flask import Flask
from loggier.integrations.flask import LoggierFlask

app = Flask(__name__)
loggier = LoggierFlask(api_key="your-api-key")
loggier.init_app(app)

FastAPI Integration

from fastapi import FastAPI
from loggier.integrations.fastapi import LoggierFastAPI

app = FastAPI()
loggier = LoggierFastAPI(api_key="your-api-key")
loggier.init_app(app)

Troubleshooting

Common Issues

1. Logs not appearing:

   • Check API key and URL
   • Ensure network connectivity
   • Check if async_mode is enabled and call flush() before exit

2. High memory usage:

   • Reduce max_batch_size
   • Increase flush_interval
   • Disable caching if not needed

3. Performance impact:

   • Use async_mode=True (default)
   • Be selective with STATUSES_TO_TRACK
   • Use track_endpoint only on critical endpoints
   • Limit body capturing (CAPTURE_REQUEST_BODY, CAPTURE_RESPONSE_BODY)

4. Django middleware error:

   • Ensure correct middleware order
   • Check LOGGIER settings in settings.py

Best Practices

1. Use context for structured logging:

   with logger.context(user_id=123):
       logger.info("User logged in")
   

2. Add custom context to Django requests:

   from loggier import _loggier_instance
   if _loggier_instance:
       _loggier_instance.context.update_global({
           "user_id": request.user.id,
           "transaction_id": "TX12345"
       })
   

3. Call flush() before application exit:

   # Ensure all logs are sent
   logger.flush()
   

4. Filter sensitive data:

   logger = Loggier(
       api_key="your-api-key",
       sensitive_fields=["password", "credit_card", "ssn"]
   )
   

5. Use decorators for performance monitoring:

   @logger.trace_function(threshold_ms=100)
   def slow_function():
       # Performance will be logged if execution exceeds 100ms