Re: bash 4.0 breaks some completion scripts

["SZEDER Gábor" <szeder@ira.uka.de>]

Hi,


On Thu, Dec 02, 2010 at 05:38:48PM -0600, Jonathan Nieder wrote:
> Jonathan Nieder wrote:
> 
> > Bash 4.0 changed the rules for completion word splitting so that
> > they are incompatible with 3.x.  I think some Breaks: will be needed
> > to ensure smooth upgrades.
> 
> Just FYI: Ah, at last some clarity!
> 
> From the bash 3 manual:
> 
>  COMP_WORDS
> 	An array variable (see Arrays below) consisting of the
> 	individual words in the current command line.  The
> 	words are split on shell metacharacters as the shell
> 	parser would separate them.  This variable is
> 	available only in shell functions invoked by the
> 	programmable completion facilities (see Programmable
> 	Completion below).
> 
> From the bash 4 manual:
> 
>  COMP_WORDS
> 	An array variable (see Arrays below) consisting of the
> 	individual words in the current command line.  The
> 	line is split into words as readline would split it,
> 	using COMP_WORDBREAKS as described above.  This
> 	variable is available only in shell functions invoked
> 	by the programmable completion facilities (see
> 	Programmable Completion below).

Oh dear, oh dear.  I've just sent a reply to your message on the git
list with basically the same findigs. [1]

> The workaround used by the bash-completion scripts is to paste words
> from COMP_WORDS together (in _get_comp_words_by_ref).
> 
> If only the NEWS file had mentioned it. :)  

Actually, the NEWS file does mention it (and Peter noted it right at
the start! [2]):

i.  The programmable completion code now uses the same set of characters as
    readline when breaking the command line into a list of words.

But it's really easy to misinterpret this statement, because there is
and there was a $COMP_WORDBREAKS, so the first thing that
comes to mind is that the set of characters in $COMP_WORDBREAKS is
changed, leaving you puzzled for weeks when you discover that the
set of characters is still the same.


Best,
Gábor


[1] http://marc.info/?l=git&m=129133327312420&w=2
[2] http://thread.gmane.org/gmane.comp.version-control.git/159516/focus=159628