Settings: list the languages alphabetically by their own name (#720)

- English stays first as the source language and fallback; the rest sort by native name with accents folded, so Čeština is under C and Magyar under M; Cyrillic and CJK names follow the Latin ones
- Lang::for_picker holds the order; the picker used the registry order, which is the order the languages were added

crates/ui-egui/src/i18n/lang.rs

@@ -275,6 +275,17 @@ impl Lang {
         LANGUAGES.iter().map(Lang)
     }
 
+    /// Every registered language in the order the Settings picker lists them: English first
+    /// (the source language and the fallback), then the rest by their own name with accents
+    /// folded, so Čeština sits under C rather than after Z. Names in other scripts follow the
+    /// Latin ones in code point order: Cyrillic, then the CJK languages. The order follows the
+    /// native name even where the picker shows a CJK language's English name instead.
+    pub fn for_picker() -> Vec<Lang> {
+        let mut rest: Vec<Lang> = Lang::all().filter(|l| l.code() != "en").collect();
+        rest.sort_by_cached_key(|l| name_sort_key(l.name()));
+        std::iter::once(Lang::EN).chain(rest).collect()
+    }
+
     pub fn name(self) -> &'static str {
         self.0.name
     }
@@ -355,3 +366,24 @@ fn chinese(tag: &str) -> Option<Lang> {
         None
     }
 }
+
+/// `name` lowercased with the Latin accents that occur in language names folded away, so a
+/// byte-wise sort reads alphabetically. Other scripts are left as they are.
+pub(super) fn name_sort_key(name: &str) -> String {
+    name.chars()
+        .flat_map(char::to_lowercase)
+        .map(|c| match c {
+            'à'..='å' | 'ā' | 'ă' | 'ą' => 'a',
+            'ç' | 'ć' | 'č' => 'c',
+            'è'..='ë' | 'ē' | 'ě' | 'ę' | 'ế' | 'ệ' => 'e',
+            'ì'..='ï' | 'ī' => 'i',
+            'ñ' | 'ń' | 'ň' => 'n',
+            'ò'..='ö' | 'ø' | 'ő' => 'o',
+            'ś' | 'š' => 's',
+            'ù'..='ü' | 'ū' | 'ů' | 'ű' => 'u',
+            'ý' | 'ÿ' => 'y',
+            'ź' | 'ż' | 'ž' => 'z',
+            c => c,
+        })
+        .collect()
+}

crates/ui-egui/src/i18n/tests.rs

@@ -4,7 +4,7 @@
 use std::collections::{BTreeSet, HashSet};
 
 use super::catalog::{Catalog, unescape};
-use super::lang::{plural_czech, plural_east_slavic, plural_polish};
+use super::lang::{name_sort_key, plural_czech, plural_east_slavic, plural_polish};
 use super::{LANGUAGES, Lang, LangInfo, lang_from_tag, message, tr, trn};
 
 /// Each catalog's rows as (line number, context, source, translation), split exactly as the
@@ -547,3 +547,18 @@ fn complete_catalogs_translate_package_formats() {
         }
     }
 }
+
+#[test]
+fn picker_lists_english_first_then_names_alphabetically() {
+    let order = Lang::for_picker();
+    assert_eq!(order.first().map(|l| l.code()), Some("en"));
+    assert_eq!(order.len(), Lang::all().count());
+    let keys: Vec<String> = order[1..].iter().map(|l| name_sort_key(l.name())).collect();
+    assert!(keys.windows(2).all(|w| w[0] <= w[1]), "{keys:?}");
+    let pos = |code: &str| order.iter().position(|l| l.code() == code).unwrap_or_else(|| panic!("{code} registered"));
+    assert!(pos("cs") < pos("de"), "Čeština sorts under C, not after Z");
+    assert!(pos("fr") < pos("hu"), "Magyar sorts under M, not under its code");
+    assert!(pos("pt-BR") < pos("ru"), "Latin names come before Cyrillic");
+    assert!(pos("ru") < pos("ja"), "Cyrillic names come before CJK");
+    assert_eq!(name_sort_key("Čeština"), "cestina");
+}

crates/ui-egui/src/settings_dialog.rs

@@ -123,7 +123,7 @@ impl LauncherUi {
             let width = ui.available_width();
             egui::ComboBox::from_id_salt("language").selected_text(Settings::language_label(&lang_code, installed)).width(width).show_ui(ui, |ui| {
                 ui.selectable_value(&mut lang_code, Settings::LANGUAGE_AUTO.to_string(), tl!("Auto"));
-                for lang in crate::i18n::Lang::all() {
+                for lang in crate::i18n::Lang::for_picker() {
                     ui.selectable_value(&mut lang_code, lang.code().to_string(), Settings::language_label(lang.code(), installed));
                 }
             });