[R6RS] End of file object

[Michael Sperber]

This seems a good candidate for a vote.  Here's the rationale from
SRFI 81:

> No distinct end of file object
>
> In R5RS, the distinct type of end of file objects is primarily for
> the benefit of read, where end of file must be denoted by an object
> that read cannot normally return as a result of parsing the
> input. However, it does not seem necessary to drag in the
> complications of this separate object into the other I/O operations,
> where #f is perfectly adequate to represent end of file.

Alternatively, I cringe every time I type:

(let ((thing (read-char p)))
  (if (eof-object? thing)
      ...
      ...))

and would rather type:

(cond
  ((read-char p)
   => (lambda (ch)
        ...))
  (else ...))

-- 
Cheers =8-} Mike
Friede, Völkerverständigung und überhaupt blabla