Re: [Patch] do not bind \340 key sequence in readline

[Benjamin Lindner]

John W. Eaton wrote:
> On  6-Nov-2007, Benjamin Lindner wrote:
>
> | Michael Goffioul wrote:
> | > Please consider the following patch. This avoids binding \340 key
> | > in readline. This key code corresponds indeed to 'à'. I modified the
> | > readline code I use for MSVC binary package, so it's no problem.
> | > However, this might have an impact on mingw version.
> | 
> | I checked with my local mingw32 build, and yes, this would be an impact
> | Unsetting the \340 key binding results in the cursor UP/DOWN keys not 
> | being functional any more.
>
> I'm surprised by this since the bindings
>
> | >   "\e[A": history-search-backward
> | > - "\340H": history-search-backward
>
> | >   "\e[B": history-search-forward
> | > - "\340P": history-search-forward
>
> should be for the feature of searching backward/forward in the history
> list when there is something already typed on the command line.  For
> simply moving up and down in the history list, the bindings should be
>
>   next-history can be found on "\C-n", "\M-OB".
>   previous-history can be found on "\C-p", "\M-OA".
>
> I found these bindings by typing
>
>   C-x k 
>
> at the Octave prompt.

I'm sorry, my comment was inprecise.
commenting out (i.e. unbinding) the \340 binding means that cursor 
UP/DOWN does no longer do a history search forward/backward. instead 
they behave as next-history/previous-history, i.e "\C-p"/"\C-n" (even if 
you typed something and then press cursor up/down) - so they are still 
functional (in the sense that pressing them results in action), but no 
longer in the intended sense (i.e. resulting in unexpected action).

> | I don't know if this is a mingw32-readline specific problem or a cmd.exe 
> | specific problem or somewhere else.
> | I have used your readline patch for compiling with mingw32, but I'm 
> | currently still at version 2.9.12. Have tehre been updates to your 
> | readline patch? (I need to get up-to-date again...)
> | 
> | We can just keep it in the mingw32 build, but this would be one 
> | difference then between msvc and mingw
>
> I think it would be best if there were no user-visible differences
> other than that .oct files are built with MinGW in one case and MSVC
> in the other.