Re: completion ignoring

Re: completion ignoring

[Stefan Monnier]

Timothy J Luoma <luomat+Lists/Zsh/users@luomat.peak.org> writes:
> # mkt[tab]
> mktargz      mktargz,v    mktargz.~1~

put the `*,v' files in a RCS subdirectory (rcs automatically put the files
there if the directory already exists)


	Stefa

PS: don't put `.' in your PATH.  After all `./mktargz' works just as well

Re: completion ignoring

[Bart Schaefer]

On May 29,  4:14pm, Vincent Lefevre wrote:
} Subject: Re: completion ignoring
}
} On Wed, May 27, 1998 at 17:21:18 +0100, Bruce Stephens wrote:
} > That's what I understood the question to be.  And yes, isn't this how
} > zsh works right now?
} 
} No:
} 
} $ benchmul[TAB]
} benchmul-sparc4  benchmul-ultra1  benchmul.o
} 
} (These 3 files are in a directory that is in my PATH.)
} 
} -rw-r--r--   1 vlefevre lip          1388 May 29 16:12 sparc/benchmul.o

On May 29,  4:36pm, Bruce Stephens wrote:
} Subject: Re: completion ignoring
}
} OK, I think that's a bug in 3.0.5 then.

It's a deficiency (?) in building the command hash table, specifically.
If you enter a full (or even relative) path to a prefix of the name of
any non-executable file and then attempt to do command completion on it,
it fails.  But if you look at the output of "hash" you'll see that every
file in every directory in $PATH has been put in the table -- and command
completion reads the hash table by default, without further testing the
file permissions on the target of the hashed name.

gen_matches_files(), which is called for things not in the PATH, does do
a stat() and look for the S_IEXEC bit.

I don't think hashtable.c has changed much in 3.1.x, so this problem is
likely still present.  However, stat()-ing every file in every directory
in $PATH is probably prohibitively slow, so this could reasonably only be
handled as a special case during completion.

Unfortunately, the code that determines whether a word matches the prefix
is completely generic -- it doesn't know that the word came from the
command hash table, or even that it's a command being completed -- and it
works by adding the word to the completion list if it matches the prefix
or by ignoring it if it doesn't match the prefix.  There's no hook for
discovering that the word *might* match and then doing an extra test on
it using any other information (such as the target of the hash).  So this
behavior is unlikely to be changed anytime soon.

Of course, you could do something like

    function xhash () {
      emulate -R zsh
      setopt localoptions
      local maybe cmd i=0
      reply=()
      hash -f
      maybe=( $(hash -m ${1}\*${2}) )
      for cmd in ${maybe##*\=}
      do
        ((++i))
	[[ -x $cmd ]] && reply=( $reply ${${maybe[$i]}%\=*} )
      done
    }

    compctl -C -aBF -K xhash -g '*(*)'

This says, attempt to complete any aliases, builtins or functions, any
executables in the hash table, and any executable files.  The `xhash'
function fills the hash table, lists everything in it that matches the
prefix and suffix of the completion, then tests the path to which each
hashes to see if it is executable before adding it to the result.

-- 
Bart Schaefer                                 Brass Lantern Enterprises
http://www.well.com/user/barts              http://www.brasslantern.com

Re: completion ignoring

[Vincent Lefevre]

On Fri, May 29, 1998 at 16:36:20 +0100, Bruce Stephens wrote:
> OK, I think that's a bug in 3.0.5 then.  It could be your
> configuration (if things work properly with "zsh -f" then you should
> look there),

I have the same problem with "zsh -f", so it seems to be a bug...

> What does the manpage zshcompctl say, under the -C option?  For 3.1.2,
> it explicitly says:
[snip]

It says the same thing (with "executable").

-- 
Vincent Lefevre <vlefevre@ens-lyon.fr> | Acorn Risc PC, StrongARM @ 202MHz
WWW: http://www.ens-lyon.fr/~vlefevre/ | 20+2MB RAM, Eagle M2, TV + Teletext
PhD st. in Computer Science, 2nd year  | Apple CD-300, SyQuest 270MB (SCSI)
-----------------------------------------------------------------------------

Re: completion ignoring

[Bruce Stephens]

Vincent Lefevre <Vincent.Lefevre@ens-lyon.fr> writes:

> $ benchmul[TAB]
> benchmul-sparc4  benchmul-ultra1  benchmul.o
> 
> (These 3 files are in a directory that is in my PATH.)
> 
> $ ll sparc/benchmul*
> -rwxr-xr-x   1 vlefevre lip         14924 May 29 16:12 sparc/benchmul-sparc4*
> -rwxr-xr-x   1 vlefevre lip         14924 May 29 13:27 sparc/benchmul-ultra1*
> -rw-r--r--   1 vlefevre lip          1388 May 29 16:12 sparc/benchmul.o

> $ echo $ZSH_VERSION 
> 3.0.5

OK, I think that's a bug in 3.0.5 then.  It could be your
configuration (if things work properly with "zsh -f" then you should
look there), but it may be something that was added since.  I only
have 3.1.2 and later (which are still beta, but which will be released
properly relatively soon, I hope), so I can't easily check 3.0.5.

What does the manpage zshcompctl say, under the -C option?  For 3.1.2,
it explicitly says:

	 -C     controls completion when the command word itself is
		   being completed.  If no compctl -C command has been
		   issued,    the  names  of  any  executable  command
		   (whether in the path or specific to the shell, such
		   as aliases or functions) are completed.

Re: completion ignoring

[Vincent Lefevre]

On Wed, May 27, 1998 at 17:21:18 +0100, Bruce Stephens wrote:
> That's what I understood the question to be.  And yes, isn't this how
> zsh works right now?  I don't remember this being changed recently, so
> I'd guess it's quite old behaviour?

No:

$ benchmul[TAB]
benchmul-sparc4  benchmul-ultra1  benchmul.o

(These 3 files are in a directory that is in my PATH.)

$ ll sparc/benchmul*
-rwxr-xr-x   1 vlefevre lip         14924 May 29 16:12 sparc/benchmul-sparc4*
-rwxr-xr-x   1 vlefevre lip         14924 May 29 13:27 sparc/benchmul-ultra1*
-rw-r--r--   1 vlefevre lip          1388 May 29 16:12 sparc/benchmul.o
$ which benchmul.o
benchmul.o not found
zsh: exit 1
$ echo $ZSH_VERSION 
3.0.5

-- 
Vincent Lefevre <vlefevre@ens-lyon.fr> | Acorn Risc PC, StrongARM @ 202MHz
WWW: http://www.ens-lyon.fr/~vlefevre/ | 20+2MB RAM, Eagle M2, TV + Teletext
PhD st. in Computer Science, 2nd year  | Apple CD-300, SyQuest 270MB (SCSI)
-----------------------------------------------------------------------------

Re: completion ignoring

[Shawn Leas]

Gee, maybe zsh should determine whether your kernel supports the a.out you
were about to try to run, or whether it thinks it's really wise that you
run that SUID script at that priority...  Or maybe we should have it check
for viruses and clip the hedges, mow the lawn, change your diapers...

-shawn
<=========== America Held Hostage ===========>
   Day 1953 for the poor and the middle class. 
   Day 1972 for the rich and the dead.
   969 days remaining in the Raw Deal.
<============================================> 

On Wed, 27 May 1998, Vincent Lefevre wrote:

> On Fri, May 22, 1998 at 09:52:32 -0700, Bart Schaefer wrote:
> > Or put the ,v in a directory that isn't in $PATH, which would make the
> > most sense anyway.  I'm not sure how the .~1~ file got created, but it
> > probably shouldn't be in the search path either.
> 
> But why does zsh take all the files from the directories that are
> in $PATH, and not only the files that can be executed by the user?
> 
> -- 
> Vincent Lefevre <vlefevre@ens-lyon.fr> | Acorn Risc PC, StrongARM @ 202MHz
> WWW: http://www.ens-lyon.fr/~vlefevre/ | 20+2MB RAM, Eagle M2, TV + Teletext
> PhD st. in Computer Science, 2nd year  | Apple CD-300, SyQuest 270MB (SCSI)
> -----------------------------------------------------------------------------
> 

Re: completion ignoring

[Bart Schaefer]

On May 27,  5:41pm, Sven Guckes wrote:
} Subject: Re: completion ignoring
}
} I think Vincent.Lefevre meant this:
} 
} When using command name expansion (correct term?)
} can the zsh check the permissions on the executable files
} and present those that are actually executable for the user?

That already happens; that's what Bruce Stevens' meant.

When you check an executable into RCS, the ,v file has write permission
removed but not execute permission:

zagzig[23] ci bleep
bleep,v  <--  bleep
initial revision: 1.1
done
zagzig[24] ls -l bleep*
-r-xr-xr-x   1 schaefer schaefer      249 May 27 09:27 bleep,v*

That's how RCS keeps track of what permissions to assign when checking
the file out again.  But zsh has no way of knowing that the ,v file is
"not really" executable.

-- 
Bart Schaefer                                 Brass Lantern Enterprises
http://www.well.com/user/barts              http://www.brasslantern.com

Re: completion ignoring

[Bruce Stephens]

Sven Guckes <guckes@math.fu-berlin.de> writes:

> Example:  Let's assume that the current dir (ie '.') is in the $PATH [*].
> Then the zsh should behave like this:
> 
> 	$ touch foo
> 	$ chmod 700 foo
> 	$ f<TAB>
> 	"foo" is shown
> 	$ chmod 600 foo
> 	$ f<TAB>
> 	"foo" is NOT shown
> 
> Is this possible?  (I hope that's not in the manual. ;-)

That's what I understood the question to be.  And yes, isn't this how
zsh works right now?  I don't remember this being changed recently, so
I'd guess it's quite old behaviour?

I just tried with "zsh -f" (3.1.2-zefram3), and it has exactly the
behaviour you list above.  Not only that, but "chmod 670 foo" still
has foo not being displayed.

Re: completion ignoring

[Sven Guckes]

Vincent Lefevre <Vincent.Lefevre@ens-lyon.fr> writes:
> But why does zsh take all the files from the directories that are
> in $PATH, and not only the files that can be executed by the user?

Quoting Bruce Stephens (b.stephens@isode.com):
> It does only consider executable files.  There's nothing wrong with a
> command ending in ~1~ or ,v.  Other than looking at the executable
> bit, how should zsh determine which files are really executable?

I think Vincent.Lefevre meant this:

When using command name expansion (correct term?)
can the zsh check the permissions on the executable files
and present those that are actually executable for the user?

Example:  Let's assume that the current dir (ie '.') is in the $PATH [*].
Then the zsh should behave like this:

	$ touch foo
	$ chmod 700 foo
	$ f<TAB>
	"foo" is shown
	$ chmod 600 foo
	$ f<TAB>
	"foo" is NOT shown

Is this possible?  (I hope that's not in the manual. ;-)

[*] Yes, I know this is a bad thing.  it's just an example, ok?  ;-)

Sven

-- 
Sven Guckes guckes@math.fu-berlin.de            using zsh-3.0.5 [961218]
ZSH HomePage: http://www.zsh.org   | Latest version:  zsh-3.1.3 [980501]
ZSH Pages:    http://www.math.fu-berlin.de/~guckes/zsh/  Tips and tricks

Re: completion ignoring

[Bruce Stephens]

Vincent Lefevre <Vincent.Lefevre@ens-lyon.fr> writes:

> But why does zsh take all the files from the directories that are
> in $PATH, and not only the files that can be executed by the user?

It does only consider executable files.  There's nothing wrong with a
command ending in ~1~ or ,v.  Other than looking at the executable
bit, how should zsh determine which files are really executable?

Re: completion ignoring

[Vincent Lefevre]

On Fri, May 22, 1998 at 09:52:32 -0700, Bart Schaefer wrote:
> Or put the ,v in a directory that isn't in $PATH, which would make the
> most sense anyway.  I'm not sure how the .~1~ file got created, but it
> probably shouldn't be in the search path either.

But why does zsh take all the files from the directories that are
in $PATH, and not only the files that can be executed by the user?

-- 
Vincent Lefevre <vlefevre@ens-lyon.fr> | Acorn Risc PC, StrongARM @ 202MHz
WWW: http://www.ens-lyon.fr/~vlefevre/ | 20+2MB RAM, Eagle M2, TV + Teletext
PhD st. in Computer Science, 2nd year  | Apple CD-300, SyQuest 270MB (SCSI)
-----------------------------------------------------------------------------

Re: completion ignoring

[Bart Schaefer]

On May 22,  5:00pm, Bruce Stephens wrote:
} Subject: Re: completion ignoring
}
} Timothy J Luoma <luomat+Lists/Zsh/users@luomat.peak.org> writes:
} 
} > Hrm..... did I do this wrong?
} > 
} > # fignore=(.o .elc \~ .wmd ,v .\~1\~)
} > # mktar[tab]
} > mktar        mktargz      mktargz,v    mktargz.~1~
} 
} That's not a file, it's a command name.  The rules are different for
} those.  You'll need to look at compctl, I'm afraid.

Or put the ,v in a directory that isn't in $PATH, which would make the
most sense anyway.  I'm not sure how the .~1~ file got created, but it
probably shouldn't be in the search path either.

-- 
Bart Schaefer                                 Brass Lantern Enterprises
http://www.well.com/user/barts              http://www.brasslantern.com

Re: completion ignoring

[Bruce Stephens]

Timothy J Luoma <luomat+Lists/Zsh/users@luomat.peak.org> writes:

> Hrm..... did I do this wrong?
> 
> # fignore=(.o .elc \~ .wmd ,v .\~1\~)
> # mktar[tab]
> mktar        mktargz      mktargz,v    mktargz.~1~

That's not a file, it's a command name.  The rules are different for
those.  You'll need to look at compctl, I'm afraid.

Re: completion ignoring

[Timothy J Luoma]

Hrm..... did I do this wrong?

# fignore=(.o .elc \~ .wmd ,v .\~1\~)
# mktar[tab]
mktar        mktargz      mktargz,v    mktargz.~1~


TjL

Re: completion ignoring

[Bruce Stephens]

Timothy J Luoma <luomat+Lists/Zsh/users@luomat.peak.org> writes:

> Is there a way to make completion ignore files that end with ,v and .~1~ ??   

Yes.

> If so, how and if not, when ;-?

See zshparam(1), under "fignore".

fignore=(.o .c~ .old .pro .tt ,v .~1~)

ought to do it.  (Change the details to taste.)

Re: completion ignoring

[Andrew Main]

Timothy J Luoma wrote:
># mkt[tab]
>mktargz      mktargz,v    mktargz.~1~
>
>Is there a way to make completion ignore files that end with ,v and .~1~ ??   
>If so, how and if not, when ;-?

fignore=(,v .\~1\~)

or, if this is for a particular command, use a ~ exclusion in a compctl
-g pattern.

-zefram

Re: completion ignoring

[Matthew Lovell]

On 22 May 1998, Timothy J Luoma writes:
 > 
 > I've recently started using Emacs RCS for some scripts.
 > 
 > The problem is that I'm getting the comma-v and the backup files when I go  
 > to do tab completion:
 > 
 > # mkt[tab]
 > mktargz      mktargz,v    mktargz.~1~
 > 
 > Is there a way to make completion ignore files that end with ,v and .~1~ ??   
 > If so, how and if not, when ;-?

Try something similar to this...

# filename completion suffixes to ignore
fignore=(.o .c~ .old .pro \~)

-- 

Matthew Lovell                       voice: (970) 898-6264 
Hewlett-Packard FSL                    fax: (970) 898-2510 
3404 E. Harmony Rd. MS A0         location: 3UR4
Fort Collins, CO 80528-9599         mailto:lovell@fc.hp.com

completion ignoring

[Timothy J Luoma]

I've recently started using Emacs RCS for some scripts.

The problem is that I'm getting the comma-v and the backup files when I go  
to do tab completion:

# mkt[tab]
mktargz      mktargz,v    mktargz.~1~

Is there a way to make completion ignore files that end with ,v and .~1~ ??   
If so, how and if not, when ;-?

Thanks

TjL