Re: fresh-line misbehavior

[Sam S. <sd...@gn...>]

> * David Tolpin <qi...@qn...g> [2005-06-05 00:28:12 +0500]:
>
>         (princ "xxx")
>         (terpri *standard-output*)
>         (princ #\Tab *standard-output*)
>         (fresh-line *standard-output*)
>         (princ "yyy")
>
>
> prints two lines in CLISP, both 2.33.2, and current CVS.
>
> xxx
> 	yyy
>
> It should print three lines
>
> xxx
>
> yyy
>
> because while #\Tab is a semi-standard character, if an implementation
> recognizes it, it is a character on the line, and the printer and not at
> the start of a new line anymore. CMUCL, Lispworks, Allegro Common Lisp,
> Open MCL and ABCL print three lines.

this problem boils down to (char-width #\tab) ==> 0
which comes from gnulib/uniwidth/width.c:uc_width 
which returns -1 because (apparently) it thinks tab is a control
character.
without UNICODE, char_width goes by graphic_char_p, 
which is required to return false for tab by
<http://www.lisp.org/HyperSpec/Body/sec_13-1-4-1.html>
and thus this behavior is NOT unicode-specific.

the fresh-line contract 
<http://www.lisp.org/HyperSpec/Body/fun_terpricm_fresh-line.html>:
  fresh-line ... outputs a newline only if the output-stream is not
  already at the start of a line. If for some reason this cannot be
  determined, then a newline is output anyway.

so, the question is: what does "at the start of a line" mean?

please reply in
https://sourceforge.net/tracker/index.php?func=detail&aid=1834193&group_id=1355&atid=101355

-- 
Sam Steingold (http://sds.podval.org/) on Fedora release 8 (Werewolf)
http://memri.org http://palestinefacts.org http://israelunderattack.slide.com
http://truepeace.org http://camera.org http://dhimmi.com http://ffii.org
Good programmers treat Microsoft products as damage and route around it.