How to sort taxonomies, taking in account accentuated letters?

Support
1

My blog is in French, and I’ve setup a tag cloud, but I got accentuated tag at the end of the list

écriture coming after zola for instance.

Is there a way to print the tag accentuated but in the correct sort order?

2

Hmm Zola doesn’t do any sorting on the tags so it might just be the serialization, maybe just tags | sort (maybe with | reverse if needed).

Si tu as du temps, https://github.com/getzola/zola/issues/768 a besoin d’avis niveau accents et autre :slight_smile:

3

I tried | sort but it stays at the end:

{% for ptag in terms | sort %}
  <span class="site-tag">
  {% set nb_pages = ptag.pages | length %}
  <a href="{{ ptag.permalink }}" style="font-size: {{ nb_pages * 4 + 80 }}%">{{ ptag.name }} ({{ ptag.pages | length }})</a>
</span>
    {% endfor %}

So we might have a bug?

(j’irais jeter un œil dès que possible :slight_smile: )

4

might be in Tera:

impl GetSortKey for String {
    fn get_sort_key(val: &Value) -> Result<Self> {
        let str: Result<&str> =
            val.as_str().ok_or_else(|| format!("expected string got {}", val).into());
        Ok(str?.to_owned())
    }
}

Maybe somekind of transformation needed here to make all accents their ASCII equivalent for sorting?

EDIT: I used https://lib.rs/crates/deunicode when implementing Jekyll’s filter slug in liquid-rust some time ago: https://github.com/cobalt-org/liquid-rust/commit/21a5be0b0f538ae67d31e5a23180f88af95df69d

5

If I want to tackle this, which package in zola I should point to my local version of Tera for experiments?

6

Probably easier to replace all of them, at least that’s what I do.

7

Ok thanks!

Another thing: on next branch, I noticed that components/rendering is not using the same version (beta.10) as the others (beta.17), is that intended?

8

Not a big deal, Cargo will use the same version since the beta versions are semver compatible

1 Like