Support EU characters in UPPER & LOWER functions

[brodycj]

A customer is interested in support for the EU characters in the UPPER function, as a quick solution to support case-insensitive string comparisons.

The LOWER function would likely be updated as well in order to maintain symmetry as much as possible.

The implementation would likely be similar to: https://www.sqlite.org/src/dir?ci=4979f138e8c8bef7&name=ext/icu

It is not yet determined whether or not to update LIKE and REGEXP.

P.S. This feature should function consistently on all supported platforms (Android, iOS, macOS, and Windows).

[brodycj]

The following EU UPPER character mappings were requested by the customer:

• ä ⇨ Ä
• à ⇨ À
• á ⇨ Á
• â ⇨ Â
• ã ⇨ Ã
• å ⇨ Å
• ǎ ⇨ Ǎ
• ą ⇨ Ą
• ă ⇨ Ă
• æ ⇨ Æ
• ā ⇨ Ā
• ç ⇨ Ç
• ć ⇨ Ć
• ĉ ⇨ Ĉ
• č ⇨ Č
• đ ⇨ Ď
• ď ⇨ Đ
• ð ⇨ Đ
• è ⇨ È
• é ⇨ É
• ê ⇨ Ê
• ë ⇨ Ë
• ě ⇨ Ě
• ę ⇨ Ę
• ė ⇨ Ė
• ē ⇨ Ē
• ĝ ⇨ Ĝ
• ģ ⇨ Ģ
• ğ ⇨ Ğ
• ĥ ⇨ Ĥ
• ì ⇨ Ì
• í ⇨ Í
• î ⇨ Î
• ï ⇨ Ï
• ı ⇨ I
• ī ⇨ Ī
• į ⇨ Į
• ĵ ⇨ Ĵ
• ķ ⇨ Ķ
• ĺ ⇨ Ĺ
• ļ ⇨ Ļ
• ł ⇨ Ł
• ľ ⇨ Ľ
• ñ ⇨ Ñ
• ń ⇨ Ń
• ň ⇨ Ň
• ņ ⇨ Ņ
• ö ⇨ Ö
• ò ⇨ Ò
• ó ⇨ Ó
• ô ⇨ Ô
• õ ⇨ Õ
• ő ⇨ Ő
• ø ⇨ Ø
• œ ⇨ Œ
• ŕ ⇨ Ŕ
• ř ⇨ Ř
• ß ⇨ ẞ
• ś ⇨ Ś
• ŝ ⇨ Ŝ
• ş ⇨ Ş
• š ⇨ Š
• ș ⇨ Ș
• ť ⇨ Ť
• ţ ⇨ Ţ
• þ ⇨ Þ
• ț ⇨ Ț
• ü ⇨ Ü
• ù ⇨ Ù
• ú ⇨ Ú
• û ⇨ Û
• ű ⇨ Ű
• ũ ⇨ Ũ
• ų ⇨ Ų
• ů ⇨ Ů
• ū ⇨ Ū
• ŵ ⇨ Ŵ
• ý ⇨ Ý
• ÿ ⇨ Ÿ
• ŷ ⇨ Ŷ
• ź ⇨ Ź
• ž ⇨ Ž
• ż ⇨ Ż

In most if not all cases, the LOWER EU character mapping would simply be the reverse of the above UPPER character mappings.

[DanielSWolf]

The conversion can never be truly symmetric because there are cases where multiple lower-case characters correspond to the same upper-case character. For example, both ď and ð will become Đ in UPPER. If the result were given to LOWER, it would have no way of knowing what the original character was.

For this reason, we plan on using only the UPPER function for comparisons. We have no need for an extended LOWER function. But it wouldn't be a problem for us if you implemented both.