Some languages use cyrillic letters that are not generally widespread across many other languages, but are a modification of a more common letter- pretty much like an accent, but they are not recognized as such in Anki. It’s probably due to those letters not having great support in general. I have been walked through using nc:, re: and w:as well as using the accent search setting in preferences but while they help with case, it doesn’t work for these (rare?) accent letters, so I was told to post here. (russian accents like ы → ы́work fine)
For a few examples:
у should find ү,ұ
н should find ң, њ, ӈ, ԩ
г should find ғ, ґ
к should find қ, ӄ
о should find ө
е already finds ё properly
I don’t know how they did it, but the Wiktionary already has it working for every letter here (except о/ө for some reason). So, for instance, I can type маґыңа or магына into the Wiktionary and it will bring me to the page мағына without issue. However, none of the 3 commands stated above could find this word in Anki unless the writing is identical. (but re: and w: can find МАҒЫНА in caps). Thank you
It’s a matter of diacritical marks, which modify a symbol, and another matter of a symbol that only looks similar, but for some reason you want to find it.
There are Russian symbols for “и” and “й,” and they’re not the same as the diacritical mark “й й”—there are two symbols that look similar, but one is simple, and the other has a diacritical mark, meaning if you omit it, it’s the letter “и” but not “й.”
So for such different and complex cases, use the search “re:[иййИЙЙ]”
I suppose that’s fair! I treated them as accents because it works like that on Wiktionary, and because their latin equivalents translate to things like ñ, ü, ū, ğ… And I add their latin equivalents, in my cards, so i can find мағына by typing magyna despite it being stored as mağyna, but I can’t do the same with the cyrillic equivalent.
I know that they are not stored as accents in the digital characters though, which is why I don’t think it’s Anki’s fault, and I guess you could argue whether or not they “should” be
If you want to enter: н should find ң, њ, ӈ, ԩ
And there it found similar ones… well, it seems a slightly different algorithm needs to be used here, I don’t even know what it is, but anything can be invented as an add-on.
In the map design itself, I remember when there were different symbols meaning the same thing, so I simply included this in the code so that it could be entered both ways. But, of course, this isn’t for searching.