Add 'title' to collection URLs

[kansaichris]

Spiritual successor to #2799.

This addresses @parkr's comment from #2799:

I'd like them to map to Liquid variables (i.e. page.name == :name, etc). Ideally, we'd have:

1. name == File.basename(path, ".*")
2. title == page.title in Liquid, sluggified

This also addresses @gjtorikian's comment from #2847:

Oh dang, I didn't know :title slugified.

Yeah, on documents would be grand.

Let me know what you think! ❤️

[parkr]

This looks great! I'd like to make the handling of slug creation a bit simpler and remove it from the url_placeholders method. Maybe sluggify(name) and sluggify(title) || sluggify(name)?

[parkr]

Thanks for doing the actual coding here though, it's very appreciated. ❤️

[gjtorikian]

Brilliant. 🎉

[parkr]

What do you think about this simplification?

if name
  name.downcase.gsub # ...
end

Or you can use unless name.nil?. IIRC, methods return nil by default. What do you think?

[parkr]

Also can we use something else besides name? It's pseudo-reserved I think.

[alfredxing]

I'm beginning to think that a sluggify function would turn out to be immensely useful throughout the rest of the project as well, and since it doesn't really have anything to do with Documents in the first place, would it be better to move it to Utils?

[gjtorikian]

I also think it should be called slugify, FWIW.

[gjtorikian]

Also, I suggest just stealing Rails method. \w is clever, but for truly "pretty" URLs it should be both lowercase and accepting of - or _.

[parkr]

👍 for moving it to Utils and for yoinking it from Rails.

[kansaichris]

I agree 💯% and will update this pull request accordingly. 😄

Incidentally, wouldn't it also be useful to expose the slugify method in a Liquid template? If so, should that wait for a separate pull request?

[parkr]

Thank you!

If you want to add it to Jekyll::Filters, go ahead 😄 I think that would be a valuable add. It can just pass to Jekyll::Utils.slugify.

[parkr]

Looking gooooooood.

[kansaichris]

Thanks for all the ✨ feedback, everyone! Let me know if you'd like me to make any more changes—I'd be particularly interested in suggestions for further unit tests. 😄

[kansaichris]

@parkr Does this look good to merge?

[parkr]

I'll take a look tonight or (more realistically) in the morning tomorrow. Sorry for the delay!

Thanks for all your contributions to Jekyll!!!!

[kansaichris]

My pleasure! 😄 Just wanted to make sure that I didn't forget anything here.

[parkr]

is this ensuring that there are no double-hyphens? i.e. no "this is my !#-fave blogs" !~> "this-is-my--fave-blogs?

[gjtorikian]

No, it ensures there's no starting or ending hyphen: "¡Hola!" !~> "hola".

[kansaichris]

What @gjtorikian said. 😄

Actually, the first regex ensures that there are no double hyphens:

string.gsub(/[^a-z0-9]+/i, '-')

Is this the desired behavior, or should we allow double hyphens in some situations?

[gjtorikian]

I've never seen a double hyphen and though "That's a good use of a double hyphen in a URL!" 😸

[parkr]

No, it ensures there's no starting or ending hyphen

Merp, yes, I meant to refer to the entire method.

the first regex ensures that there are no double hyphens

Great! Then we're good to go.

[parkr]

LGTM excepting the above comment about double-hyphens.

[parkr]

Thanks for all your hard work, @kansaichris! I think I'll try to wrap up a 2.4.0 release of Jekyll sometime this weekend or on Monday.

[gjtorikian]

@parkr Humble ask if we can work on tagging #2870 into 2.4.0 too? We're working on some internal shtuff that depends on these two pieces. Would prefer to talk offline because of sensitive internally stuff about this. 😀

[kansaichris]

🎉