Unicode slugs are not properly slugified due to javascript limitations

[Sævar Öfjörð Magnússon]

When using unicode slugs, and the slug contains a unicode character and a word from the removelist in urlify.js right after it (e.g. the letter "a"), the unicode character is detected as a word boundary in javascript and the text that matches the removelist is stripped from the slug...
The removelist contains the following words:

var removelist = [
"a", "an", "as", "at", "before", "but", "by", "for", "from", "is",
"in", "into", "like", "of", "off", "on", "onto", "per", "since",
"than", "the", "this", "that", "to", "up", "via", "with"
];

And when I try to slugify this text: "Kaupa miða"
The result becomes: "kaupa-mið"

This can be tested using the following line in console:
URLify("Kaupa miða", 255, true)

(only when urlify.js is loaded, of course)

[Claude Paroz]

The solution could be to check if the string contains any non-ASCII char and simply skip the removelist removal, as the language is probably not English in that case.

[Sævar Öfjörð Magnússon]

I've created a pull request that implements Claude's suggestion: ​https://github.com/django/django/pull/9219

[Tim Graham <timograham@…>]

In f90be0a8:

Fixed #28688 -- Made admin's URLify.js skip removal of English words if non-ASCII chars are present.

[Tim Graham <timograham@…>]

In 8b9a163:

Refs #28688 -- Updated a selenium test for admin's URLify.js change.

English words aren't removed if non-ASCII chars are present.