Re: [Bug-readline] Cannot use M-<key> bindings in incremental search

[Chet Ramey]

On 8/29/10 1:03 AM, Davor Cubranic wrote:
> On August 28, 2010 04:55:41 pm you wrote:
>>> Could this be because the lookup of the key in _rl_isearch_dispatch
>>> (isearch.c:340) does not match for bindings involving Meta because
>>> the input character 'c' does not map to an ISFUNC in the current
>>> _rl_keymap?
>>
>> It is, but changing this is a problem.  I can easily modify the code
>> so that it handles keymap[key].type == ISKMAP and behaves
>> appropriately, but it introduces an incompatibility.  ESC is one of
>> the characters that terminates an incremental search; making this
>> change will break that.  I'm reluctant to change the behavior as a
>> result.
> 
> How about doing this check only if 'key' is not in 'isearch-
> terminators'? I should have added that I removed ESC from my list of 
> terminators, precisely for this reason.

That's the way I had done my experimental implementation, with a
few other changes.


> Or, even better, move the isearch-terminators check as early as 
> possible. Right now, it doesn't look like a few special characters (C-
> G/W/Y) will work as they're treated specially around lines 350-355, and 
> the terminator check is at line 362. That way, the whole customization 
> works more consistently, and no keys are treated specially.

That doesn't quite work, because of the desire to have escape-prefixed
sequences like the arrow keys work by default.

I think we'll try allowing different key sequences in readline-6.2, and
see how it works.

Chet

-- 
``The lyf so short, the craft so long to lerne.'' - Chaucer
                 ``Ars longa, vita brevis'' - Hippocrates
Chet Ramey, ITS, CWRU    address@hidden    http://cnswww.cns.cwru.edu/~chet/