Logo Questions Linux Laravel Mysql Ubuntu Git Menu
 

Are uppercase utf8 characters always the same number of bytes as their lowercase variants?

Obviously it is true for the latin alphabet. But I'm asking this in a conceptual sense, across languages and the Unicode spec.

Practically this came up for comparing two strings. If you already know they aren't the same number of bytes—across all languages—can you consider that enough of a guarantee that they are not differently "cased" versions of the same string?

like image 358
HostileFork says dont trust SE Avatar asked Feb 09 '13 23:02

HostileFork says dont trust SE


Video Answer


2 Answers

No.

Consider U+0069 "i" which has the octet value 69 in UTF-8. In the uppercase form U+0130 "İ" this code point forms the UTF-8 sequence C4 B0.

Obligatory note: case is locale-sensitive.

like image 179
McDowell Avatar answered Sep 20 '22 15:09

McDowell


There is no principle or invariant in the Unicode standard that guarantees this. I would be particularly concerned about accented capitals, where there may be a mismatch between precomposition and non-precomposition across cases. However, I can't cite an example of a problem for you.

like image 25
bmargulies Avatar answered Sep 21 '22 15:09

bmargulies