Add {{type:nc:…}} – ignores combining characters when comparing

[twwn]

This allows to add :nc to type fields, causing Anki to ignore all combining characters so as to not fault their absence in the user input as errors.

It improves one's learning experience with languages such as Hebrew whose use of diacritics (nikud) is almost entirely limited to reading them (in class/references), with nobody writing them (very cumbersome anyway) in everyday use for ordinary words.

Example before / after:

The approach basically consists of:

• Strip given & correct of their combining characters
• Use that for tokenizeComparison
• "Map" the stripped correct to its original
• Restore the combining characters in the correction line with that "mapping"

It's either-or, meaning any combining characters in the input/given are discarded as well (so no partial comparison).

Notable otherwise: one functional change affecting the non-:nc path and two clean-ups

• If no user input (given) was entered, simply return the field/correct straight away with minimal formatting.
• Removed an unused correct() parameter.
• Skip the redundant tokenizeComparison for the non-:nc path when given already matches correct.

[dae]

Hi rathsky,

No objections to the idea itself, but the type-in-the-answer code is likely going to move into TypeScript in the future, so this would probably be better done after that change is made - will leave this open until then.

Regarding the specifics of this implementation, perhaps simply stripping all combining characters if the user has used nc: would suffice? It would make for a simpler implementation if you don't need to preserve the original text verbatim.

[twwn]

It would make it simpler, but while the diacritics are an obstacle to typing the words, they are still essential to their learning. So removing them entirely would require duplicating the field, which is unsightly and a waste of space.

[dae]

Cleaning up old PRs - please open a new one if you'd like to continue this.