Lower-case Latin variants hide upper diacritics about diacritics.net

[thomasgalliker]

This library works with replacement mappings: Diacritics character í is mapped to i, character Í is mapped to I, and so on. There is no general mechanism that maps between diacritics and non-diactritics characters. Can you report all characters (and their normalized equivalents) please?

[anth12]

For Turkish there are 4 I's:

The lowercase version of İ is i
The lowercase version of I is ı

The mappings to "remove diacritics" would be:
İ -> I
ı -> i

However with the current code, simply adding the mapping wouldn't solve due to the lower case conversion as İ would be converted to i & therefore not replaced.

[thomasgalliker]

Hmm, I'm lost somehow. So, if I add these two mappings, it would not be enough to solve the problem?
İ -> I
ı -> i

[anth12]

There are mappings for the characters, but the issue lies in DiacriticsMapper#56.

Say I invoke DiacriticsMapper.RemoveDiacritics("İngiltere"), at L56 input = "İngiltere" & source = "ingiltere" which leads to the original input being returned as the while checks the lowercase source for instances of any diacritic character.

[thomasgalliker]

Ups, I somehow forgot to come back to this issue. Do you have a proposal on how to solve it? Is it because line 56 does only compare index positions ignoring case sensitivity?

[thomasgalliker]

@anth12 can you please check this nuget pre-release:
https://www.nuget.org/packages/Diacritics/2.1.19240.1-pre

[thomasgalliker]

@anth12 did you have a chance to install and test the latest nuget pre-release? I want to make sure your issue is solved before I close it.

[thomasgalliker]

Seems like you have your own diacritics library now 🤩 I guess I can close this issue then.