Hacker News Reader

The Why is "Å" !== "Å" !== "Å"? section still strikes me as wrong. The strings are equal even when the representations differ.

View on HN

They are logically equal (that is, they represent the same text in an abstract way), but computing this equality in practice is expensive, because you first need to normalize the strings then compare.

Most languages, when comparing strings, skip the normalization and just compare string bytes as is (or, if the string is interned, compare just the pointer)

You can easily do the comparison dynamically with checking for combining marks, and then do the proper lookup. No need to normalize everything, or even store the normalized variant. Though in a filesystem or username lookup you would only store it normalized.

I just not sure why they put in the "Angstrom symbol" to begin with. If you do, then why isn't the "meter symbol" (m) also represented?

Fortunately, it seems like it's marked as deprecated: https://en.wikipedia.org/wiki/Angstrom#Symbol

> I just not sure why they put in the "Angstrom symbol" to begin with.

Frequently, the answer to this is "some obscure character set had this as a distinct symbol." In this case, blame the Japanese: https://en.wikipedia.org/wiki/JIS_X_0208

Which is why there's an 'mm' and 'cm' and other random symbols: https://www.compart.com/en/unicode/block/U+3300

Reply on news.ycombinator.com