[Conkeror] conkeror-xr: string input with history?
Mark Eichin
eichin at gmail.com
Mon Nov 5 22:55:21 PST 2007
err, not quite that simple; had to chop out the pass through
get_url_or_webjump() :-) but that was a start, at least.
On 11/6/07, Mark Eichin <eichin at gmail.com> wrote:
> Hmm, looking at emacs' (interactive), it appears that it's simpler
> just because it's lisp, so it only needs "simple string" and "eval and
> expect a simple string." Though that model might work here too, even
> if it means using a list of strings; it lets you put one is-function
> check up front...
>
> (For my particular case, it turns out that to get simple history, I
> can just use "url_or_webjump" instead of "s", since there's no actual
> url-sanity check :-)
>
> Hmm, emacs also kind of cheats in that (completing-read) isn't a part
> of (interactive) at all - it just goes and gets the answer, and you're
> trying to combine them more closely (which I think I favor, but it
> does explain the list-of-functions a little better.)
>
>
>
> On 11/5/07, John J Foerch <jjfoerch at earthlink.net> wrote:
> > "Mark Eichin" <eichin at gmail.com> writes:
> > > I need an interactive() mode that lets me prompt for a string with
> > > history. I used to do this with miniBufferComplete; while [["s",
> > > "prompt: "]] is working to get the input at all, m-p doesn't show any
> > > prior values.
> > >
> > > Should I be adding something to interactive_methods? And if so...
> > > what I really want is something to let me input a list of single-word
> > > tag names, with completion (and history.) Should I just start with
> > > what 'b' or 'f' do now and experiment?
> > >
> > > --
> > > _Mark_ <eichin at thok.org> <eichin at gmail.com>
> >
> > Hi Mark,
> >
> > interactive_methods.s should be improved to allow for history to be
> > provided by the caller. Right now I'm not satisfied with
> > interactive's way of passing arguments to interactive methods. You
> > may notice that in interactive.js, some interactive methods require
> > their arguments to be passed as functions, while others, like `s' will
> > take a simple string prompt. Also, arguments to interactive methods
> > are always passed by position, which makes it hard to provide a
> > consistent interface. If you have any ideas for how to overcome these
> > problems, or even just a patch to get a quick fix for what you want to
> > do right now, I would be glad to know and assist.
> >
> > --John
> >
> > _______________________________________________
> > Conkeror mailing list
> > Conkeror at mozdev.org
> > http://mozdev.org/mailman/listinfo/conkeror
> >
>
>
> --
> _Mark_ <eichin at thok.org> <eichin at gmail.com>
>
--
_Mark_ <eichin at thok.org> <eichin at gmail.com>
More information about the Conkeror
mailing list