cadadr / gemini-scripts

My Gemini scripts.

Göktuğ’s Gemini Scripts

This is a catch-all repository for my scripts related to the new Gemini protocol.

This work, including this document, is licensed under the Cooperative Non-Violent Public License, version 6 or newer, whose text is available in the LICENSE file in this repository, or online at https://git.pixie.town/thufie/CNPL. In case the local and online versions differ, the more recent version applies.

g2html.awk

g2html.awk converts Gemini files to HTML files.

$ gawk -f g2html.awk input.gmi > output.html

If either or both of header.html or footer.html are found in the working directory, they are appended and prepended, respectively, to output.html.

If you want to use other files as the header and/or the footer, you can set the variables header and footer at the command line:

$ gawk -v header=/path/to/header.html -v footer=/path/to/footer.html \
    -f g2html.awk input.gmi > output.html

In order to convert a tree of Gemini files, you can use find(1) as follows:

find gemini -name '*.gmi' -exec sh -c 'gawk -v header=templates/header.html -v footer=templates/footer.html -f scripts/g2html.awk {} > html/wormhole/$(basename -s .gmi {}).html' \;

Admittedly, that’s a gross command, so you might want to convert this to a for loop and put it in a script instead.

g2ra.py

g2ra.py is an RSS/Atom proxy for subscribable Gemini pages. Uses Flask and Ignition libraries.

g2ra.py can also be used as a command line app to generate a feed, given a gemlog index file.

Major caveat: currently does not handle Gemini redirects.

Command line frontend

Generate static feeds from local files

This could come in handy for you if you are auto-generating a HTML version of your gemlog. The operation is simple:

% g2ra.py static -u my.url -t atom -a "Author" -p path/to/index.gmi

See g2ra.py static --help for more info. The Flask frontend is more lenient when it comes to default values for parameters, command line frontend has a couple more required ones.

Run test server

You can run a test server as follows:

% g2ra.py flask

This will start a local CGI server you can use.

Proxy frontend

I may one day deploy this on some public server (go ahead if you want to do so yourself, patches welcome!), but as it is now, you are expected to deploy this script on some server yourself. It should be fairly easy to do using usual Flask deployment options, and below are instructions on how to set it up with Gunicorn and virtualenvs as a local service.

First of all, create a virtual environment:

% mkdir g2ra-deploy
% cd g2ra-deploy
% python3 -m venv venv

Activate the virtual environment and install dependencies:

% source venv/bin/activate
% pip3 install gunicorn flask ignition-gemini

Test the script is running alright:

% # copy g2ra.py here
% gunicorn -b 127.0.0.1:1961 'g2ra:flask_app("__main__")'
% curl 'localhost:1961?url=cadadr.space/blag.gmi'

Write a simple script you run at login:

% cd ..
% cp -r g2ra-deploy ~/local/g2ra-deploy
% cat <<EOF > ~/bin/start-g2ra.sh
> #!/bin/sh
> cd ~/local/g2ra-deploy
> ./venv/bin/python -m gunicorn -b 127.0.0.1:1961 'g2ra:flask_app("__main__")'
> EOF
%

Here on you can use any of ~/.config/autostart, user systemd units, cron, .xinitrc, etc to run this script.

Another likely option is to run this as a (Fast)CGI script using a local installation of some web server. I haven't set that up yet, but following flask instruction should be enough. I might extend this tutorial later with relevant info.

The usage of g2ra.py is simple. You will use a combination of the the following query parameters:

• url: mandatory, the gemini page to convert. g2ra does not do redirects yet, so you need the full path to a gemini page.

• type: optional, can be either of atom or rss. The former is the default, and causes the page to be converted to an Atom feed. The latter will lead to an RSS feed being generated.

• author: Author field for Atom scripts' mandatory <author> fields. The default value is Unknown Author Jr..

• ttl: Time to live attribute for RSS channels (the <ttl> tag), defaults to 1800.

gemini-scripts's Issues

make g2ra.py work on command line too, read stdin & write to stdout

g2ra.py: abort() if type query parameter is unexpected

currently it raises an exception which will result in a 500 error which is not what we want. it should probably be a 400 instead.

g2ra.py: --pretty causes xml parser to fail with --type rss