Debating whether to use an existing HTML pre-processor or just forge
on ahead with Python. So far, using beautifulSoup, I've managed to
generate a table of contents automatically, and of course clean up
my indenting. It does generate #ids for headings if they don't have
their own, but it doesn't work as well for the little blog section
on my index. This is all about my Neocities, by the way.

What I want out of a template:
```
<blogpost>
<htime datetime="2022-10-23" />
<p>Here's a bunch of text</p>
</blogpost>
```

to render as
```
<section id="2022-10-23">
  <h3><time datetime="2022-10-23">October 23, 2022</time></h3>
  <!-- or h4, h5, depends on the page -->
  <a href="table-of-contents">Back to top</a>
  <!-- hidden with CSS, only for keyboard navigation -->
  <p>Here's a bunch of text</p>
</section>
```

Additionally,
```
<codeblock>
  <summary>Graphing code</summary>
  code goes here
</codeblock>
```

to render as
```
<details>
  <summary>Graphing code</summary>
  <pre><code>
code goes here
  </code></pre>
</details>
```

Is there a semantic reason to use <code>pre</code>? I'm pretty sure
I can just slap a class on the <code>code</code> element and be done
with it, but if there's a semantic reason I probably shouldn't.