logstack-py 1.0.2

A Python SDK for the Logstack logging platform

Logstack Python SDK

A Python SDK for the Logstack logging platform.

Installation

pip install logstack-py

The import name is still logstack (e.g. from logstack import LogStackClient).

v1.0.2

• capture_logging default on. Automatically forwards stdlib logging calls with source: "python-logging". Original handlers (console, etc.) are preserved.
• Lowers root logger level when needed so info / debug records reach the capture handler.
• Skips re-entrant capture from the internal Logstack logger.

v1.0.1

• Fix batch-flush deadlock when batch_size is reached
• Normalize API URL (strips redundant /v1 suffix)
• Optional on_error callback; accepts HTTP 201 from ingest API
• create_fastapi_middleware(client) for proper FastAPI/Starlette integration

Usage

Basic Usage

from logstack import LogStackClient

# Create a client
# capture_logging=True (default): any logging.getLogger(...).info / error etc.
# is automatically captured and sent with source="python-logging".
# Your normal logging configuration (console handlers etc.) continues to work.
client = LogStackClient(
    api_key="your-api-key",
    environment="production",
)

# Explicit structured logs
client.info("Application started", metadata={"version": "1.0.0"})
client.error("Database connection failed", metadata={"error": "connection refused"})

# Flush and close
client.close()

Automatic stdlib logging capture

By default capture_logging=True. All Python logging calls across your app (including libraries) are forwarded:

import logging
from logstack import LogStackClient

client = LogStackClient(api_key="...")  # capture_logging=True by default

log = logging.getLogger(__name__)
log.info("User action", extra={"user_id": 123})   # <- captured automatically
log.error("Failed to process")                    # <- captured as error level

The original log records still go to any handlers you have configured (StreamHandler, etc.).

Disable with capture_logging=False if you only want explicit client.xxx() calls.

Using Context Manager

from logstack import LogStackClient

with LogStackClient(api_key="your-api-key") as client:
    client.info("Application started")
    client.error("Something went wrong")

Django Integration

# settings.py
MIDDLEWARE = [
    # ...
    'logstack.middleware.DjangoMiddleware',
    # ...
]

# Or with custom client
MIDDLEWARE = [
    # ...
    'logstack.middleware.DjangoMiddleware',
    # ...
]

# In your code
from logstack import DjangoMiddleware, LogStackClient

client = LogStackClient(api_key="your-api-key")
middleware = DjangoMiddleware(get_response, client=client)

FastAPI Integration

from fastapi import FastAPI
from logstack import LogStackClient, FastAPIMiddleware

app = FastAPI()
client = LogStackClient(api_key="your-api-key")
FastAPIMiddleware(app, client=client)

API Reference

LogStackClient(api_key, ..., capture_logging=True)

Create a new Logstack client.

info(message, metadata=None)

Send an info level log.

debug(message, metadata=None)

Send a debug level log.

warn(message, metadata=None)

Send a warn level log.

error(message, metadata=None)

Send an error level log.

critical(message, metadata=None)

Send a critical level log.

fatal(message, metadata=None)

Send a fatal level log and flush immediately.

flush()

Manually flush the batch of logs.

close()

Close the client and flush any pending logs.

License

MIT