Re: lynx-dev Re: [BUG] Editing in unicode

[Thomas E. Dickey]

On Mon, 30 Jun 2003, Ilya Zakharevich wrote:

> On Mon, Jun 30, 2003 at 06:15:45AM -0700, Ilya Zakharevich wrote:
> > Is it possible to use Lynx with utf-8 displays *at all*?  I go to
> > google, customize it to use Russian; reload the search page to get the
> > new display.
> >
> > Now select any word in xterm, and try to paste it into the search
>                      ^^^^^^^^^
>
>   I meant: suppose that lynx is running in `xterm -u8'; select any
>   Russian word shown in the "Russian" google search page.
>
> > field.  I get junk [empty squares].  [Something different (?!!!)
> > happens if I try to paste a word shown by lynx in a different page: I
> > get ######## instead of empty squares.]
>
> More details: the search field shows junk; but if I press Tab to go to
> the "Search" button (labeled in Russian, of course ;-), the search
> field becomes correct.  Now pressing the Search button activates a
> correct search.
>
> But now select the search term repeated on the *new* page (same word
> as before!); press Left to go back to the main search page.  Try
> pasting this word into the search field again.  I get #####.
>
> I'm completely puzzled...

What you're describing isn't unexpected (unless I'm misinterpreting based
on my own expectations).  7-bit ASCII text is not a problem.  8-bit codes
(128-255) for Unicode aren't 8-bit UTF-8 codes (they use more than one
byte).  Since lynx doesn't know how to read UTF-8 from the keyboard, it
displays the input as garbage.

The past couple of weeks (interleaved with other tasks) I've been
rewriting dialog's inputstr.c to use wget_wch() and related functions
to allow it to read UTF-8.  The same approach (not a rewrite) could be
used for lynx.

-- 
T.E.Dickey <address@hidden>
http://invisible-island.net
ftp://invisible-island.net

; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to address@hidden