Add `nomark` for striping accents, `samemark` is counterintuitive and slow for that purpose.

[bbkr]

Stripping accents to get base characters is very common operation in text indexing and searching. Raku allows to do it through:

"mówić".samemark( "a" )

There are two issues with that method:

• For someone not knowing documented remaining characters in $string receive the same mark/accent as the last character in $pattern behavior it is confusing and just looks weird. It's just arbitrairly chosen magic "repeat last character" WAT behavior.
• Underlying implementation takes unnecessary steps because it must parse $pattern and extract decomposed grapheme codepoints of last character. With all the hassle of monitoring $pattern length at the same time.

My proposal is to add easy to use, optimized method for stripping accents:

say "mówić".nomark()   # mowic

One that will preferably avoid reallocating whole string splitting/joining if there are no marks to strip. Because usually this method will be called blindly on any given input - with or without marks. As Raku is more commonly used with LLMs this may be good addition to language.

[coke]

regexes have not only samemark but alsoignoremark

https://docs.raku.org/syntax/%3Aignoremark

In case we wanted to follow that naming convention. (I realize it's not a 100% match)

[bbkr]

@coke: I think this is early design inconsistency, probably dated back to Apocalypses.

• samemark is adjective (which?)
• ignoremark is noun (what to do?), should be ignoreDmark

That's why I'm not big fan of copying this mixed naming. I was also thinking about basemark name.

[lizmat]

https://github.com/rakudo/rakudo/pull/5562 contains an implementation.