diff options
Diffstat (limited to 'README.md')
| -rw-r--r-- | README.md | 140 |
1 files changed, 0 insertions, 140 deletions
diff --git a/README.md b/README.md deleted file mode 100644 index 03e640c3..00000000 --- a/README.md +++ /dev/null @@ -1,140 +0,0 @@ -# gemini_blog - -This repo contains my personal blog, hosted on Gemini. -The blog is basically a static Gemini capsule. - -This repo also contains scripts I used to migrate my Wordpress blog to Gemini. - -## Running the blog - -The `build.py` script renders a static blog using `.gmi` files in a `content` directory. -The static blog capsule goes to the `build/gmi` directory: - -``` -$ ./build.py -``` - -To preview the site locally, you can use https://github.com/michael-lazar/jetforce, that serves static `.gmi` files and can create automatic certificates. -Note that Jetforce does not fully support Python 3.6. -You can use `pipx` to install Jetforce: - -``` -$ pipx install jetforce -``` - -Serve your built blog: - -``` -$ jetforce --dir build/gmi/ --host 0.0.0.0 -``` - -To preview, use a client like https://github.com/makeworld-the-better-one/amfora : - -``` -$ amfora localhost:1965 -``` - -To run the blog in production, I am currently using https://github.com/mbrubeck/agate in a `screen` session: - -``` -$ ./agate --content build/gmi/ -``` - -You will need `cert.pem` and `key.rsa` in the same directory for the proper host, as Gemini mandates the use of TLS. - -I use `xinetd` from a host with a public IP to proxy to the host running `agate`: - -``` -service gemini -{ - disable = no - type = UNLISTED - socket_type = stream - protocol = tcp - wait = no - redirect = $HOST_RUNNING_AGATE 1965 - bind = 0.0.0.0 - port = 1965 - user = nobody -} -``` - -To create an HTML proxy of the capsule: - -``` -$ docker build -t kineto -f Dockerfile.proxy . -$ docker run --rm -p 5000:5000 kineto -``` - -Or to preview the HTML rendering of a local running instance: - -``` -$ docker run --rm --network host kineto /app/kineto gemini://localhost -``` - -To generate an RSS feed: - -``` -$ ( cd gemini2rss/ ; poetry run python gemini2rss.py https://alex.corcoles.net 10 "El blog es mío" "https://alex.corcoles.net" Alex alex@corcoles.net ; ) >/tmp/feed.rss -``` - -## Migrating a Wordpress blog to Gemini - -Note: only posts are migrated. -Notably, images, categories, tags, comments, and pages are not migrated. - -First, create an export from the Wordpress admin interface. -This will generate an XML file containing all posts and comments. - -Use https://github.com/lonekorean/wordpress-export-to-markdown to generate Markdown files out of the blog's posts. -This tool is the simplest tool I could find that can convert the Wordpress export into something simpler to convert to Gemini. - -To run the tool, you need a most recent version of Node.js than what my current workstation has. -If you cannot run `npx wordpress-export-to-markdown`, then you can use https://github.com/alexpdp7/cmdocker to run a newer version of Node.js using Docker: - -``` -$ cmdocker add-wrapper node_15 /bin/bash node:15 -$ node_15 -``` - -Run `wordpress-export-to-markdown` to convert your Wordpress archive into Markdown files in a directory named `md`: - -``` -$ npx wordpress-export-to-markdown --input $YOUR_WORDPRESS_XML \ - --year-folders true \ - --month-folders true \ - --post-folders=false \ - --wizard false \ - --output md -``` - -Use `md2gemini` to convert the Markdown files into `.gmi` files. -I use https://github.com/pipxproject/pipx to install Python tools in isolated user virtualenvs: - -``` -$ pipx install md2gemini -``` - -Perform the conversion: - -``` -$ find md -name '*.md' -execdir md2gemini -l at-end -w '{}' ';' -$ find md -name "*.md" -delete -``` - -The converted `.gmi` files need some fixing, as the title, posted date, categories are not in a easy-to-use format. -The `fix.py` script takes all `.gmi` files in the `md` directory, "fixes" them into a new `content` directory: - -``` -$ ./fix.py -``` - -## More stuff - -``` -$ find content/ -name '*.gmi' -execdir sed -i 's|http[^ ]*watch?v=|https://farside.link/invidious/watch?v=|g' {} \; -``` - -``` -$ { cat alex.corcoles.net_access.log alex.corcoles.net_access.log.1 ; zcat alex.corcoles.net_access.log.*gz ; } | cut -d \" -f 2 | grep GET | cut -d " " -f 2 | grep -E -e '^/[0-9]{4}/[0-9]{2}/.+' | grep -v '/feed/$' | sort | uniq -c | sort -n -``` |