Not all locales use single-character exponent separator

TIL: some locales in CLDR do not have a single-character exponent separator.

• 25 Arabic-based locales use "اس",
• 11 Arabic-based locales use "×۱۰^" (albeit some of them are marked AnyScript),
• 8 Latin-script locales use "×10^" (Scandinavia, Lithuania and Estonia),
• 3 Latin-script locales use "·10^" (Northern Sami).

QLocale has been using the first character of each of these as the exponent separator in relevant locales (and I've never seen it reported as a bug).
Naturally, we should also check the other "single-character" data in the API actually are for all locales; a check in the generator script can take care of that.
(For reference, seven locales use "e" (C, English (Australia, Slovenia), Greek, Slovak, Slovenian), 569 use "E" at least officially; but that includes several English locales (US, GB) in which "e" is at least sometimes used; and Ukrainian uses "Е", which may look like "E" but is Cyrillic, hinting that some other Cyrillic scripts might, in practice, also use this. These 577 do use single-character exponents; with the above, QLocaleData has 624 locales in CLDR v41.)