search

scala / scala3

The Scala 3 compiler, also known as Dotty.
https://dotty.epfl.ch
Apache License 2.0
5.89k stars 1.06k forks source link

scaladoc: javascript #21637

Open Quafadas opened 2 months ago

Quafadas commented 2 months ago

Compiler version

3.3.3

Minimized code

filename : js.md

<h1>Hello JS</h1>

<script type="text/javascript">
  console.log("boo")
</script>

In the "_docs" directory, of a static site, which will be part of a scaladoc static site.

Output

If you generate the site, serve it locally and navigate to the URL inhabited by this file (e.g. localhost:3000/docs/js.html), you'll see that the browser console has "boo" printed.

If you navigate to a different page on the sidebar (e.g. localhost:3000/docs/notjs.html), then navigate to localhost:3000/docs/js.html via the sidebar, no message will be printed. In other words, javascript doesn't fire if you arrive via the internal # routing.

This breaks my expectations, and I think it is awkward, because it works "the first time", but not subsequently, which is an awkward user / developer experience.

Expectation

I would expect javascript to run when navigating to the page. It is possible, that this is a legitimate design choice, but it feels odd, because it does work as expected, the first time.

I will note that I think there are legitimate uses for being able to run js on documentation pages. The examples which I'm interested in are;

And I'm sure there are more. I'm reporting this as a bug, perhaps in ignorance - it could be intended behaviour.

Culprit

A discord discussion led us to believe the culprit is this; https://discord.com/channels/632150470000902164/632150470000902166/1288100519029641289

image

Quafadas commented 2 months ago

Thinking about this, if it is a bug, I think it's actually kind of a nasty one because the behaviour isn't obvious and it sucks a lot of time figuring out where the problem really lay.

I think this is the second hour or three spent checking my own stuff before starting to question scaladoc itself, and I ended up lucky to get some heavyweight help finding that loadPageState function.