This is a single-author, mostly command line driven blog that renders Markdown from your favorite editor to a server side endpoint. It is primarily written to serve my yeti-factory needs, which is to say that it mostly exists to let me play with Clojure, and perhaps occasionally write a blog post.
- Programmer "friendly". Written in Lisp with what I hope is sane organization.
- Write your posts in Markdown, have them rendered to Github flavored markdown, and then get syntax highlighted code blocks.
- Ring compatible, so you can use existing Ring middleware to wrap your requests.
- Template driven with a layout style somewhat similar to Rails. Uses StringTemplate.
- Fast, thanks to Netty.
- Command-line driven, so you can use your favorite
$EDITORto write posts.
- Automatically uploads local assets to S3 if they are referenced in the outgoing post.
- Written in short spurts over the past two weeks, so you know the software quality is high
- No tests, because laziness
- Disqus integration! Google Analytics integration! Are these features! I don't know!
- RSS support!
- Better organization, ala
- Ensure Heroku support.
- Better caching header defaults.
- Environment specific behavior, like template loading.
- Maybe? not use environment variables for config everywhere.
- Remove yetifactory name, come up with a better name
- Remove my specific blog styling, keep in a separate branch. Maybe.
grep -ri 'TODO' .
- A 500 page, i don't expect it will see much use tho
- Pagination, probably for when I write > 20 blog posts. Probably a hard problem in Lisp tho, amirite eh eh
- Namespace auto uploaded images such that identical files don't conflict from post to post...or do they.
The command line driver is available in
bin/live, and may be configured in
a shell file avaliable in
$HOME/.blog. See the configuration section for
Originally taken from clojure-netty, now me playing around w/ different concepts in clojure to render webpages.
# Server side setup, set via dokku ENV, heroku env set, or similarp export USERNAME="the-username-for-editing-content" export PASSWORD="the-password-to-edit-content" export DB_HOST="the-database-host" export DB_USER="the-database-user" export DB_PASSWORD="the-database-password" export DB_NAME="the-database-name" # Disqus can be enabled by export DISQUS_NAME="your-disqus-name" # Google Analytics can be enabled by export GOOGLE_ANALYTICS_ID="your-ga-id" # Client side setup goes in .blog, can go in working directory or $HOME export USERNAME="the-username-for-editing-content" export PASSWORD="the-password-for-editing-content" export EDITOR="your-editor-command-mayhaps subl -w" export BLOGHOST="http://domain-of-your-blog.com" # If you want to do automatic file uploading to S3, you can use the # included upload_image script. You'll need the following command line # tools installed: xmlstarlet, markdown, ruby # # Once you've done that, add the following to your client side script: # export UPLOAD_SCRIPT="./bin/upload_image" export EC2_ACCESS_KEY="YOUR-EC2-ACCESS_KEY-FOR-S3" export EC2_SECRET_KEY="YOUR-EC2-SECRET_KEY-FOR-S3" # # If you want to write your own image uploader, it has to: # Accept a filename on STDIN, AND # Output "LOCAL_PATH:REMOTE_PATH" on STDOUT # Once you've done that, just use the path to your image uploader # in UPLOAD_SCRIPT as opposed to ./bin/upload_image.
You'll need to create a database using the schema in
db/schema.sql. By default
this creates a database named "blog", change beforehand if you need it to be
Once you've got your database created, you'll need to ensure the correct environment
variables are set for your config. Locally you'll likely just want to change
the values in
project.clj, which are just the defaults; in production, you'll
want to change via whatever mechanism your server has environment variables set.
lein deps lein run
will run the server. By default content is available on port
To manage content, you'll want to use
bin/live in the
will list available blog posts in CSV format
will open up
$EDITOR with a new post, and when
$EDITOR exits with
will post the content to your blog.
bin/live edit SLUG
will open up
$EDITOR with an existing post, and save it to the blog
$EDITOR closes with
bin/live destroy SLUG
will delete a post.
Notes on Deployment
I've been testing this on Dokku, which uses the Heroku Clojure buildpack. I expect that this /will/ work on Heroku, but you'll need to set the following env vars:
export DB_PROTOCOL="postgresql" export DB_DRIVER="org.postgresql.Driver"
in order to work with Postgres. You'll also need to edit project.clj to include the Postgres driver. That is assuming you're using Postgres on Heroku; other database users should edit the above envars to suit their specific needs.