python-fragment 0.4.0

Reverse engineered Fragment API that mimics the official webclient

python-fragment

An unofficial Python library for interacting with Fragment. The library scrapes Fragment's HTML pages and exposes a clean, fully typed interface.

Features

• Both synchronous and asynchronous clients
• Fast HTML parsing powered by selectolax
• Fully typed - all returned data is described with TypedDict schemas
• Automatic retries on network failures with exponential back-off

Installation

pip install python-fragment

Requirements: Python 3.9+

Quick Start

Async client

import asyncio
import fragment

async def main():
  async with fragment.AsyncClient() as client:
    # Search usernames currently on auction
    usernames = await client.search_usernames(
      query="crypto",
      filter="auction",
      sort="price_desc"
    )
    for u in usernames[:3]:
      print(u)
      # {
      #   'username': 'crypto',
      #   'status': 'auction',
      #   'value': 15000.0,
      #   'datetime': '2024-08-01T12:00:00+00:00',
      #   'is_resale': False
      # }

asyncio.run(main())

Sync client

import fragment

with fragment.Client() as client:
  # Detailed info for a specific username
  info = client.username_info("durov")
  print(info["status"])
  print(info["bid_history"])
  print(info["ownership_history"])

Proxy Usage

Both clients forward **request_kwargs directly to the underlying HTTP library, so proxies are configured per-call.

Async client (aiohttp)

Pass a single proxy URL string via the proxy keyword argument:

import asyncio
import fragment

PROXY = "http://{user}:{password}@{host}:{port}"

async def main():
  async with fragment.AsyncClient() as client:
    print(await client.username_info("crypto", proxy=PROXY))

asyncio.run(main())

Sync client (requests)

Pass a proxies dict that maps URL schemes to the proxy URL:

import fragment

PROXY = "http://{user}:{password}@{host}:{port}"

with fragment.Client() as client:
  print(client.username_info("crypto", proxies={"http": PROXY, "https": PROXY}))

Error Handling

import fragment
from fragment.errors import FragmentHTTPError, ParserError


async with fragment.AsyncClient() as client:
  try:
    info = await client.username_info("someusername")
  except FragmentHTTPError as e:
    print(f"HTTP error: {e}")
  except ParserError as e:
    print(f"Failed to parse Fragment HTML: {e}")

FragmentHTTPError - raised when the server returns a non-OK response.

ParserError - raised when the Fragment page structure has changed and data cannot be extracted.