you can use character.isWhitespace() to check this rather than comparing to our own hardcoded values

Here you're duplicating the production algorithm, meaning there's no way your test can fail.

yeah I know, this is for all the other tests that are not exactly testing the obfuscation, is just that we are asserting everything in those. For all the obfuscation cases I added the 3 tests below.

You should also have a test in the form

val chunk1 = randomText
val chunk2 = randomText
val spacer = randomWhitespace // forge.forge.aWhitespaceString()
val input = chunk1 + spacer + chunk2
val expectedOutput = ("x" * chunk1.length) + spacer + ("x" * chunk2.length)

with this (I don't know if it exist in our repo

operator fun String.times(i: Int) {
    val builder = StringBuilder(i * length)
    repeat(i) {
        builder.append(this)
    }
}

Good point

let's add TODO maybe here, saying it works only for ASCII. Also character can be non-printable Control character, but I guess we can ignore it.

I just realized that I need to use another method to support all unicode characters the way they are specifying in their comments.

this won't work properly for Unicode still. For example, let's take the following emoji. It has a code \u1f600, so it already exceeds char size (which values go only up to \uffff, because it can fit only 16 bits), so this single symbol consists of the 2 char -> "😀".length will give 2.

But in reality we cannot threat these chars independently, because first is just a high surrogate and second is the low surrogate. This is how UTF-16 encoding is actually working - it will read the stream of char of length 16 bits and decide if next char is a lower surrogate or independent symbol.

Yes I imagine this is the way it should work but I think this will not be a problem. The only issue would be if one of this emojis will contain one of the non printable characters in their composition. In that case we will brake the text but I am pretty sure Character.isWhitespace() handles that somehow or maybe these whitespace characters are not used in emojis.

My point is that in case of emoji we will have xx instead of single x, not sure if this is what we want to achieve (I guess Character.isWhitespace() will return false for any surrogate).

Yeah...I don't think this should be an issue as long as we obfuscate the texts. It will be a corner case anyway. I am going to add this explanation there though to have this in mind.

I won't say it is a corner case, because there are plenty of symbols with above \uffff, and if we think of emojis we may find a lot of them in messenger apps, for example, and obfuscation may break formatting. Not a blocker for me, but let's acknowledge this.

it seems the default repeat should do the job?

why do we need times call if it just copies the same?

to have a bigger text. I was following @xgouchet suggestion, not sure that this is what he wanted but I think is a bit overkill also. @xgouchet do you think this is still needed ?

the times was more to generate a string with just 'x' of a given size, here it doesn't make sense

char doesn't cover all the Unicode characters. It covers BMP plain Unicode characters, but it doesn't cover Supplementary plain characters (above 0xFFFF). When we call character.code we just cast from char to int, but to get code of the symbol from supplementary plain we need to merge 2 chars into single int.

It is not maskString, but obfuscate. Same for other tests.