GNU bug report logs - #17559
`initials' does not complete underscore delimited input

Previous Next

To reply to this bug, email your comments to 17559 AT debbugs.gnu.org.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Andreas Politz <politza <at> hochschule-trier.de>
To: bug-gnu-emacs <at> gnu.org
Subject: 24.3.50;
 Partial completion does not complete underscore delimited input 
Date: Fri, 23 May 2014 16:47:51 +0200

The documentation for the customizable variable
`completion-pcm-word-delimiters' states:
``"A string of characters treated as word delimiters for completion.''

The default value contains _ , yet it does not seem to work, i.e.

(let ((completion-styles '(partial)))
  (completion-initials-all-completions "fdh" '("friss_die_haelfte") nil 0))
=> nil
(let ((completion-styles '(partial)))
  (completion-initials-all-completions "fdh" '("friss-die-haelfte") nil 0))
=> ("friss-die-haelfte" . 0)

It would be nice for c-mode, if this could be made working.

Grüße,

A. Politz

Message #8 received at 17559 <at> debbugs.gnu.org (full text, mbox):

From: Stefan Monnier <monnier <at> iro.umontreal.ca>
To: Andreas Politz <politza <at> hochschule-trier.de>
Cc: 17559 <at> debbugs.gnu.org
Subject: Re: bug#17559: 24.3.50;
 Partial completion does not complete underscore delimited input
Date: Fri, 23 May 2014 11:47:20 -0400

retitle 17559 `initials' does not complete underscore delimited input
severity 17559 wishlist
thanks

> The documentation for the customizable variable
> `completion-pcm-word-delimiters' states:
> ``"A string of characters treated as word delimiters for completion.''
> The default value contains _ , yet it does not seem to work, i.e.

This variable just means that "a_b_c" can be completed by
`partial-completion' style to "alonzo_bob_church" and indeed, that
works, AFAICT.

> (let ((completion-styles '(partial)))
>   (completion-initials-all-completions "fdh" '("friss_die_haelfte") nil 0))

[ completion-styles is used to decide which completion-*-all-completions
  to call, but completion-initials-all-completions should be unaffected by
  completion-styles (and indeed, it isn't, and `partial' is not a valid
  completion style, AFAIK). ]

> => nil

Indeed, completion-initials-all-completions hard-codes "-" as the
separator (you can see it in completion-initials-expand).


        Stefan

Message #15 received at 17559 <at> debbugs.gnu.org (full text, mbox):

From: Andreas Politz <politza <at> hochschule-trier.de>
To: Stefan Monnier <monnier <at> iro.umontreal.ca>
Cc: 17559 <at> debbugs.gnu.org
Subject: Re: bug#17559: 24.3.50;
 Partial completion does not complete underscore delimited input
Date: Fri, 23 May 2014 19:57:56 +0200

Stefan Monnier <monnier <at> iro.umontreal.ca> writes:

> retitle 17559 `initials' does not complete underscore delimited input
> severity 17559 wishlist
> thanks
>

> [...] "a_b_c" can be completed by
> `partial-completion' style to "alonzo_bob_church" [...]
>

Yes, you're right.

Andreas

Message #18 received at 17559 <at> debbugs.gnu.org (full text, mbox):

From: Andreas Politz <politza <at> hochschule-trier.de>
To: Stefan Monnier <monnier <at> iro.umontreal.ca>
Cc: 17559 <at> debbugs.gnu.org
Subject: Re: bug#17559: 24.3.50;
 Partial completion does not complete underscore delimited input
Date: Fri, 23 May 2014 20:19:14 +0200

[Message part 1 (text/plain, inline)]

Stefan Monnier <monnier <at> iro.umontreal.ca> writes:

> severity 17559 wishlist

Would something like the following be sufficient ?

[vc.diff (text/x-diff, inline)]

=== modified file 'lisp/minibuffer.el'
*** lisp/minibuffer.el	2013-10-04 02:12:16 +0000
--- lisp/minibuffer.el	2014-05-23 18:14:44 +0000
***************
*** 3187,3193 ****
                  (string-match completion-pcm--delim-wild-regex str
                                (car bounds)))
        (if (zerop (car bounds))
!           (mapconcat 'string str "-")
          ;; If there's a boundary, it's trickier.  The main use-case
          ;; we consider here is file-name completion.  We'd like
          ;; to expand ~/eee to ~/e/e/e and /eee to /e/e/e.
--- 3187,3200 ----
                  (string-match completion-pcm--delim-wild-regex str
                                (car bounds)))
        (if (zerop (car bounds))
!           (let ((separator
!                  (or (completion-metadata-get
!                       (completion-metadata str table pred)
!                       'completion-initials-separator)
!                      (plist-get completion-extra-properties
!                                 :completion-initials-separator)
!                      "-")))
!             (mapconcat 'string str separator))
          ;; If there's a boundary, it's trickier.  The main use-case
          ;; we consider here is file-name completion.  We'd like
          ;; to expand ~/eee to ~/e/e/e and /eee to /e/e/e.

[Message part 3 (text/plain, inline)]

This appears to work, but I have little clue about the underlying
partial completion matching algorithm.

Andreas

Message #21 received at 17559 <at> debbugs.gnu.org (full text, mbox):

From: Stefan Monnier <monnier <at> iro.umontreal.ca>
To: Andreas Politz <politza <at> hochschule-trier.de>
Cc: 17559 <at> debbugs.gnu.org
Subject: Re: bug#17559: 24.3.50;
 Partial completion does not complete underscore delimited input
Date: Thu, 19 Jun 2014 17:41:32 -0400

> Would something like the following be sufficient ?
[...]
> !           (let ((separator
> !                  (or (completion-metadata-get
> !                       (completion-metadata str table pred)
> !                       'completion-initials-separator)
> !                      (plist-get completion-extra-properties
> !                                 :completion-initials-separator)
> !                      "-")))

It could work, tho it requires changing all the corresponding completion
tables to provide the other "filling" when needed.
I think this should be fixable/fixed without needing to touch the
completion tables.

And if we add such `completion-initials-separator' properties, we'll be
stuck with them "for ever", so I'd rather not go there if there's
a chance we might not need them.


        Stefan

Message #24 received at 17559 <at> debbugs.gnu.org (full text, mbox):

From: Andreas Politz <politza <at> hochschule-trier.de>
To: 17559 <at> debbugs.gnu.org
Subject: Re: bug#17559: 24.3.50;
 Partial completion does not complete underscore delimited input
Date: Wed, 31 Aug 2016 21:27:22 +0200

Coming back to this...

What do you think, where should the information regarding the string to use as
a separator for the initials style come from, if it's not hard-coded ?

Andreas

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.