Re: Proposed change to `make-readline-port'

[Neil Jerram]

>>>>> "Neil" == Neil Jerram <address@hidden> writes:

    Neil> Unfortunately, not quite, I think.  Only the body of
    Neil> `make-line-buffered-input-port' knows when a character has
    Neil> been read out of its internal buffer (`read-string').

That said, perhaps a better way of addressing this problem would be to
support `char-ready?' for soft ports.  Then the application logic
could be:

    (if (char-ready? port)
        (prepare to use continuation prompt)
        (prepare to use new read prompt))
    (read)

Currently, char-ready? on a soft port always returns #t.  But I can't
see a fundamental reason why this should always be so.  Perhaps all
that is needed here is a Scheme wrapper for
scm_set_port_input_waiting?

Regards,
        Neil