rememberTypewriterReveal

Reveals text a character at a time, at charsPerSecond, instead of showing it all at once.

Built for streamed content such as an assistant reply that grows text over several recompositions. When text grows, the reveal continues from where it is and catches up; when text is not an extension of what is already shown, the reveal restarts from the beginning. Surrogate pairs (most emoji) are revealed whole.

The full text is returned at once when the system reduces motion (rememberReduceMotionEnabled). The returned State is the revealed prefix, rendered however the caller likes. Screen readers read whatever is on screen, so give the finished message, not each intermediate prefix, to any live region.

Parameters