This post is a bit of an oddity since I had originally written it when I (briefly) returned to Hugo for my personal blog. I have since decided that I didn’t blog frequently enough to feel like maintaining that, but I figure the content could still be useful regardless.
By default, Hugo will create posts at a URL like:
Since I didn’t really want to end up with hundreds or even thousands of posts in a single directory, though, I decided that I wanted to create yearly sub-directories such as:
I did this mostly for my own sake. While the site was small, if I kept using it for a while then the posts directory in my git repository would end up being fairly massive. Along with being ugly, that just makes it more difficult for me to do things like:
- Find the relevant file.
- Name things uniquely.
There were just a couple of things I needed to do to make this work. First, in my
config.toml file I added a list for
mainSections = ["2022", "2023"]
Next I manually created a
2022 and a
2023 directory in
content/posts and moved my Markdown files from the root of the
posts directory to the appropriate yearly directory. Locally testing this via...
… showed that everything looked good. A commit to the repository later, and everything was live on the new site. It’s definitely worth mentioning, just in case it isn’t obvious, that this will break any existing links since my post URLs now contain the year. For me, this was desirable. Given that the site isn’t particularly large, I figured it wasn’t a big deal.
I had been curious about the process for creating new posts since I would previously just do:
hugo new posts/file-name.md
After the change, I could just specify the full path:
hugo new content/posts/2023/file-name.md
Removing Public Directory From Origin
One mistake I had been making was that I had committed my
/public directory to my repository, so the compiled content of the site was included in GitHub. That was fairly useless since I had a GitHub Action to compile the site when a new commit is pushed; my
/public directory in origin was just a lot of static content that wasn't being used for anything.
I removed the directory from origin via the trusty:
git rm -r --cache public/*
Note that if you are okay to delete the local content as well, the
--cache flag can be omitted. I next added that directory to my
.gitignore file so that any subsequent content which may end up there wouldn't accidentally be included when staging files. My initial thought was that I would still want to keep compiling the site locally to review changes, new posts, etc. prior to pushing to GitHub. However, I realized afterward that's not quite right since running
hugo server will compile the content into memory, which is what allows it to render new changes on the fly without re-running
hugo after a file has been updated and saved.
Another nice benefit to this is that, after updating the site, adding a new post, etc. locally, it’s a lot easier to tell what has changed with:
Prior to this, I’d end up with at least a few dozen files that were new, updated, etc. in
/public. It was simple enough to sort through, but it was still more effort than was necessary. Without running
hugo, the only items which show up when I need to check what files have been modified are the ones I manually modified.
Now my repo’s origin is a bit slimmer and doesn’t contain wasted files. Likewise, I can remove running
hugo from my workflow entirely.