On Fri, Sep 03, 2021 at 04:51:06PM +0200, Ingo Schwarze wrote: > Hi Theo, > > as you sent it, your patch is misleading since our manual page describes > two features that are not required by POSIX. > > So i propose the larger patch shown below instead. > > While here, clarify what "null-terminated" means. It matters for > this page in particular because the page talks about both NUL bytes > and NULL pointers in different places. > > Also, failing to state that the original input string is modified > feels like a serious omission to me. By the way, these replacements > are indeed required by POSIX. The strsep(3) below SEE ALSO provides > a rather weak hint, but i regard that as insufficient. > > OK?
Thanks, this looks good to me. ok
