Add Additional Filter

[pnewell]

I was wondering if there is some way to add additional filters (specifically non visible ASCII character (backspace, etc)?

[dzcpy]

I'm going to work on it. What I'm thinking is to provide an option which you can use your own Unicode -> Latin mappings. Something like:

transliterate('こんにちわ、世界！\u0008', { charmap: { 8 /* \u0008 backspace */: '⌫' });

This is much better than using replace option, since it uses a regex to match and replace the entire source string, splitting the source string into multiple segments and then calling transliterate in a recursive ways to transliterate each segments. If using a Unicode number to Latin mapping it's much more efficient.

At the first step, it may only support simple Unicode -> Latin character map. But later on in v2 I'd implement a plugin system, so for some specific language which cannot be 1-1 mapped can implement their own logic, only when the source string contains the characters in that language. For example, Chinese, Japanese and Thai are not working properly without analyzing each whole words, instead of single characters. Then, since I'm not familiar with the transliteration rules from a lot of languages, other users might be able to contribute their own transliteration rules.

Ultimately, if I have time, I'd like to make it possible to transliterate from 1 language to another, e.g. from English to Russian, from Spanish to Japanese etc.. There's a tone of workload though and requires much professional knowledge.

How do you guys think about it?

[dzcpy]

With v2 you can do it like this:

transliterate.setData({
  '\u0008': '⌫'
});
transliterate('こんにちわ、世界！\u0008');

[dzcpy]

I'll close it for now. If you have more questions, please comment.