🚀 go-pugleaf

The documentation for search-forward makes no explicit mention of what
happens if the repeat count is zero.  Whether or not the documentation
is changed to mention this case, I think it would be useful to change
the behavior.  The present behavior is mysterious; it seems like an
unforeseen consequence of the way more usual cases are implemented.

What seems to be happening is this:  if the cursor is near enough the
beginning of the buffer, it stays put and search-forward returns nil.
Otherwise, it moves backward some distance which seems to have nothing
to do with the contents of the buffer, and search-forward returns t.

What I think should happen with a zero repeat count is that
search-forward should leave the cursor where it is and return t,
whether or not the search pattern occurs anywhere in the buffer.  It
seems a conditional early return in search.c:search_command is a
suitable way of doing this.  I don't have diffs for that, but I'm
willing to make some, test them, and send them in if there's any need
for that.

This is more than a philosophical question to me; I was forced by this
mysterious behavior to complicate some elisp code I was writing.