partial completion and switch-to-buffer

partial completion and switch-to-buffer

[Tobias Gerdin]

Greetings,

With the advent of partial-completion-mode in Emacs 23, I was wondering
if there isn't any way to use this mechanism with switch-to-buffer ("C-x b")?

Switch-to-buffer still seems to do only prefix-matching.

I am aware of iswitch-mode and ido-mode, which does substring matching
(which is basically what I want), but find them too obtrusive. I do not
want any completion to be done unless I say so (by pressing <tab>).

I'm using Emacs CVS from 2009-07-30 (i.e. pretty much 23.1).

-Tobias

RE: partial completion and switch-to-buffer

[Drew Adams]

> I am aware of iswitch-mode and ido-mode, which does substring matching
> (which is basically what I want), but find them too obtrusive. I do not
> want any completion to be done unless I say so (by pressing <tab>).

Icicles does no completion until you say so (by pressing TAB), by default.

http://www.emacswiki.org/cgi-bin/wiki/Icicles


----

Details -

* To get substring completion instead of prefix completion, use S-TAB.

* Or, if you want to use TAB for both prefix and substring completion, then set
option `icicle-cycling-respects-completion-mode' to t. (Use TAB / S-TAB to
change the completion mode: prefix / substring.)

* S-TAB actually does regexp completion, by default. To get only substring
completion, set option `icicle-regexp-quote-flag' to non-nil. Or use `C-`' in
the minibuffer to toggle it.

* When buffer *Completions* is showing, the completion candidates are updated as
you type (incremental completion), by default. To turn this off, set option
`icicle-incremental-completion-flag' to nil. Or use `C-#' in the minibuffer to
toggle it.

Re: partial completion and switch-to-buffer

[Tobias Gerdin]

"Drew Adams" <drew.adams@oracle.com> writes:

>> I am aware of iswitch-mode and ido-mode, which does substring matching
>> (which is basically what I want), but find them too obtrusive. I do not
>> want any completion to be done unless I say so (by pressing <tab>).
>
> Icicles does no completion until you say so (by pressing TAB), by default.

I am aware of Icicles, but I would prefer not to have to rely on
an external package for this simple matter.  But thanks for the thorough
explanation.

-Tobias

comment-padding type ?

[Pierre Lorenzon]

Hi All,

comment-padding documentation is as follows :

Padding string that `comment-region' puts between comment chars and text.
Can also be an integer which will be automatically turned into a string
of the corresponding number of spaces.

But in function comment-padright we find this piece of code :

    (let ((s (match-string 1 str))	;actual string
	  (lpad (substring str 0 (match-beginning 1))) ;left padding
	  (rpad (concat (substring str (match-end 1)) ;original right padding
			(substring comment-padding ;additional right padding
				   (min (- (match-end 0) (match-end 1))
					(length comment-padding)))))

where comment-padding is actually supposed to be a sequence and
not an integer. Actually in one mode comment-padding was set to
1 and I got an error when trying to execute
comment-padright. When setting the value to " " everything goes
well ! 

In my opinion the comment-padding documentation is
abusive. Either it should be changed or a conversion mechanism
should be added.

Regards 

Pierre

Re: comment-padding type ?

[Stefan Monnier]

> But in function comment-padright we find this piece of code :

>     (let ((s (match-string 1 str))	;actual string
> 	  (lpad (substring str 0 (match-beginning 1))) ;left padding
> 	  (rpad (concat (substring str (match-end 1)) ;original right padding
> 			(substring comment-padding ;additional right padding
> 				   (min (- (match-end 0) (match-end 1))
> 					(length comment-padding)))))

> where comment-padding is actually supposed to be a sequence and
> not an integer. Actually in one mode comment-padding was set to
> 1 and I got an error when trying to execute
> comment-padright. When setting the value to " " everything goes
> well ! 

> In my opinion the comment-padding documentation is
> abusive. Either it should be changed or a conversion mechanism
> should be added.

There is a conversion mechanism.  It's in comment-normalize-vars which
you failed to call (and whose importance I obviously fail to make
clear, in the doc).


        Stefan

Re: comment-padding type ?

[Pierre Lorenzon]

Hi !

From: Stefan Monnier <monnier@iro.umontreal.ca>
Subject: Re: comment-padding type ?
Date: Wed, 26 Aug 2009 11:06:46 -0400

>> But in function comment-padright we find this piece of code :
> 
>>     (let ((s (match-string 1 str))	;actual string
>> 	  (lpad (substring str 0 (match-beginning 1))) ;left padding
>> 	  (rpad (concat (substring str (match-end 1)) ;original right padding
>> 			(substring comment-padding ;additional right padding
>> 				   (min (- (match-end 0) (match-end 1))
>> 					(length comment-padding)))))
> 
>> where comment-padding is actually supposed to be a sequence and
>> not an integer. Actually in one mode comment-padding was set to
>> 1 and I got an error when trying to execute
>> comment-padright. When setting the value to " " everything goes
>> well ! 
> 
>> In my opinion the comment-padding documentation is
>> abusive. Either it should be changed or a conversion mechanism
>> should be added.
> 
> There is a conversion mechanism.  It's in comment-normalize-vars which
> you failed to call (and whose importance I obviously fail to make
> clear, in the doc).

  OK ! it works now. 

  Thanks !

  Pierre


> 
> 
>         Stefan
>