Sort lines correct alphabetic sorting with diacritics in Romanian

[akaleeroy]

Issue Type: Bug

The correct order when it comes to Romanian diacritics would be a, ă, â, b, c ... i, î ... s, ș, t, ț, u, v ...

Alphabetical order - Wikipedia

But this is the result of Sort Lines Ascending on these lines:

Sălaj
Satu Mare
Sibiu
Suceava

Sălaj should come after Satu Mare

FWIW Sublime Text gets it worse, considering diacritics as continuing after the main latin alphabet: Sălaj comes after Suceava

VS Code version: Code 1.32.3 (a3db5be, 2019-03-14T23:43:35.476Z)
OS version: Windows_NT x64 6.1.7601

[akaleeroy]

On the other hand it matches up with the sorting from LibreOffice Calc.

[isidorn]

Sorting where? Are you talking about the explorer?

[akaleeroy]

Sort Lines Ascending in the Command Palette.

[alexdima]

This is actually more complicated than I thought...

The reason is that different locales have different sort orders for the same characters. See a full example at https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Collator

function letterSort(lang, letters) {
  letters.sort(new Intl.Collator(lang).compare);
  return letters;
}

console.log(letterSort('de', ['a','z','ä']));
// expected output: Array ["a", "ä", "z"]

console.log(letterSort('sv', ['a','z','ä']));
// expected output: Array ["a", "z", "ä"]

So there is no absolute order of letters, ordering letters/strings only makes sense in a given locale. We are currently using String.localeCompare. I believe that one falls to the locale defined in the OS, and on your machine I think that is English.

Testing the sorting:

function sort(lang, letters) {
  letters.sort(new Intl.Collator(lang).compare);
  return letters;
}

console.log(sort('en', ["Satu Mare", "Sălaj", "Sibiu", "Suceava"]));
// ==> ["Sălaj", "Satu Mare", "Sibiu", "Suceava"]

console.log(sort('ro', ["Satu Mare", "Sălaj", "Sibiu", "Suceava"]));
// ==> ["Satu Mare", "Sălaj", "Sibiu", "Suceava"]

So under English the sort order is indeed ["Sălaj", "Satu Mare", "Sibiu", "Suceava"], while under Romanian it is ["Satu Mare", "Sălaj", "Sibiu", "Suceava"]

I will change our code to instantiate a Intl.Collator, which is better than String.localeCompare from a performance point of view according to what I learned from MDN, but I think the only way to get the sort order you want is to change your OS locale to Romanian.