I see, thanks for your input. I'm happy for this to be closed now.

Thanks for your response.

The main idea behind this so to increase readability/maintainability. I think strchrnul/strspn convey the intention of the code better than a random loop. But I see why you wouldn't want to introduce a change for small things like that.
The performance implications are a bonus at best.

Hi,

I'm writing a program analysis that tries to find and replace some loops with str* functions. I'm trying to see if these replacements are a useful refactoring or the loops are intentional. I noticed that bash has some loops that can be replaced, so I wrote two patches that replace them.

The first strchrnul one replaces a loop at general.c:825 with strchrnul. I think that's more readable and faster on platforms where libc has fast versions of strchrnul.

The second replaces 11 loops across multiple files with calls to strspn. I hard coded the whitespace characters, because I couldn't find a good place to put a macro. I'm happy to change it, if you suggest a place for the constants.

I would also be interested in understanding why the loops are there in the first place or why you would like to keep them.

Thanks!