Re: problems with 4.3.4 and Tru64

[Peter Stephenson <pws@csr.com>]

Phil Pennock wrote:
> > > If these display fine, then multibyte is working and parts of zle are
> > > working; the problem will be in input processing.  If these don't
> > > display, then the problem's not in the input processing.
> > 
> >  Touche, that works!
> > 
> >  Now, how to fix input processing?-)

The fact they display doesn't tell you very much.  You also need to
check that once you've got them on the command line they behave like a
single character.  The easiest way is just to move the cursor and check
it doesn't go to far (which it will if doesn't recognise the two bytes
as a single character).

If that's really working, input processing is quite involved to check.
The key function for this is getrestchar() in Src/Zle/zle_main.c.  This
should recognise the first byte is an incomplete character and the
second byte completes it.  I don't think this an easy way to avoid
debugging the insides.

> What I have noticed is that the configure tests for multibyte changed
> between 4.3.2 and 4.3.4.  It might be helpful for those who _can_ debug
> these if you can dig out config.log and config.h from both the 4.3.2 and
> 4.3.4 builds and put them on a webserver, posting the URLs here.  At a
> _very_ brief look, it seems that there was a move to checking for
> existence of the standardised API for widechar handling.

I think we've already gone through this by showing that the multibyte
option is on by default.

-- 
Peter Stephenson <pws@csr.com>                  Software Engineer
CSR PLC, Churchill House, Cambridge Business Park, Cowley Road
Cambridge, CB4 0WZ, UK                          Tel: +44 (0)1223 692070


To access the latest news from CSR copy this link into a web browser:  http://www.csr.com/email_sig.php

To get further information regarding CSR, please visit our Investor Relations page at http://ir.csr.com/csr/about/overview