Ignition: Python3 Gemini Protocol Client Transport Library

Ignition is a simple but powerful transport library for Python3 clients using the recently designed Gemini protocol. This project intends to implement all of the transport specifications (sections 1-4) of the Gemini protocol and provide an easy-to-use interface, so as to act as a building block in a larger application.

If you're building a Python3 application that uses Gemini, Ignition is your gateway to the stars, in very much the same way that requests is for HTTP and gopherlib is for Gopher.

In order to provide a best-in-class interface, this library does not implement the other parts of a typical client (including user interface and/or command line interface), and instead focuses on providing a robust programmatic API interface. This project also assumes that different user interfaces will have different requirements for their display of text/gemini files (.gmi), and/or other mime-types, and as such considers this portion of the specification beyond the scope of this project.

In addition, in order to provide stability and simplicity, minimal third-party dependencies are required for Ignition.

Project Status

Ignition is no longer being updated.

Installation

⚠ Ignition supports Python versions 3.7 - 3.12.

Ignition can be installed via PIP. You should install it in alignment with your current development process; if you do not have a build process yet, I recommend you install within a virtual environment

pip install ignition-gemini

If you prefer to install from source, you can clone and install the repository:

git clone https://github.com/cbrews/ignition.git
cd ignition
pip install .

Simple Usage

The most basic usage of Ignition allows the user to create a request and get a response back from a remote Gemini capsule:

import ignition

# Fetch capsule content
response = ignition.request('//geminiprotocol.net')

# Get status from remote capsule
print(response.status)

# Get response information from remote capsule
print(response.data())

source

In all cases, Ignition assumes that the specified endpoint and protocol will respond over the Gemini protocol, so even if you provide a different protocol or port, it will assume that the endpoint is a Gemini capsule.

Key Features

✅ Ignition supports the following features:

Basic request/response connectivity to a Gemini-enabled server.
Basic URL parsing mechanics to allow for specifying protocol, host, port, path, and query params, as per RFC-3986
Optional referer URL handling. Ignition allows the user to pass a path & referer URL and can construct the new path, to simplifying the resolution of links on a Gemini capsule page.
Basic decoding of body responses on successful (20) response from Gemini servers.
Trust-on-first-use certificate verification handling scheme using key signatures.
Fully-featured response objects for each response type.
Standardized & robust, human-readable error management.
Custom error handling for networking failure cases beyond the scope of the protocol.

❌ The following Gemini features will not be supported by Ignition:

Behavioral processing/handling of specific response types from Gemini capsules, including:
- Generation of client certificates & automatic resubmission.
- Automatic redirection following on 3x responses.
Advanced body response rendering and/or display of text/gemini mime types.
Command line or GUI interface.
Advanced session & history management.
Support for other protocols.
Non-verified certificate scheme
Improved TOFU scenarios

Advanced Usage

More advanced request usage:

import ignition

response = ignition.request('/software', referer='//geminiprotocol.net:1965')

print("Got back response %s from %s" % (response.status, response.url))
# Got back a response 20 from gemini://geminiprotocol.net/software

if not response.success():
  print("There was an error on the response.")
else:
  print(response.data())

Passing a referer:

import ignition

response1 = ignition.request('//geminiprotocol.net')
response2 = ignition.request('software', referer=response1.url)

print(response2)

source

More advanced response validation:

import ignition

url = '//geminiprotocol.net'
response = ignition.request(url)

if response.is_a(ignition.SuccessResponse):
  print('Success!')
  print(response.data())

elif response.is_a(ignition.InputResponse):
  print('Needs additional input: %s' % (response.data()))

elif response.is_a(ignition.RedirectResponse):
  print('Received response, redirect to: %s' % (response.data()))

elif response.is_a(ignition.TempFailureResponse):
  print('Error from server: %s' % (response.data()))

elif response.is_a(ignition.PermFailureResponse):
  print('Error from server: %s' % (response.data()))

elif response.is_a(ignition.ClientCertRequiredResponse):
  print('Client certificate required. %s' % (response.data()))

elif response.is_a(ignition.ErrorResponse):
  print('There was an error on the request: %s' % (response.data()))

source

Finally, the module exposes DEBUG level logging via standard python capabilities. If you are having trouble with the requests, enable debug-level logging with:

import logging
logging.basicConfig(level=logging.DEBUG)

API Documentation

Full API documentation for Ignition is available here.

Developers

Ignition is no longer accepting contributions. Please feel free to fork this repository.

License

Ignition is licensed under Mozilla Public License 2.0.

Thank you

solderpunk for leading the design of the Gemini protocol, without which this project would not have been possible.
Sean Conman for writing the Gemini torture tests, which were instrumental in initial client testing.
Michael Lazar for his work on Jetforce, which helped testing along the way.

🔭 Happy exploring!