Thanks, that makes sense. I guess this is somewhat an argument to keep
the first hunk too, it's more of a statement that v->end is < zlen,
Although, now that the else if is gone, the same is expressed by the last test, i.e., the first hunk kind of becomes doubly redundant.
Another reason I removed it is because it's not present in the equivalent code for arrays in
setarrvalue. In
workers/54348, I unify the code for strings and arrays. If you keep the first hunk, I might add an equivalent one for arrays in that patch.
Philippe