PATCH: _man copes with : delimited manpath

PATCH: _man copes with : delimited manpath

[Adam Spiers]

On my system, `manpath' generates colon-delimited paths, rather than
space-delimited ones.  This patch solves that problem.  It also fixes
what I think is a typo, but someone should check this.

Index: Completion/User/_man
===================================================================
RCS file: /usr/local/cvsroot/zsh/Completion/User/_man,v
retrieving revision 1.1.1.4
diff -u -r1.1.1.4 _man
--- Completion/User/_man	1999/09/21 12:17:47	1.1.1.4
+++ Completion/User/_man	1999/09/27 01:14:03
@@ -14,12 +14,18 @@
   approx="(#a${_comp_correct})"
 fi
 
-(( $#manpath )) || manpath=$(manpath 2>/dev/null) ||
+if (( ! $#manpath )); then
+  local mp=$(manpath 2>/dev/null)
+  [[ "$mp" == *:* ]] && mp=( ${(s.:.)mp} )
+  manpath=( $mp )
+fi
+
+(( $#manpath )) ||
   manpath=( /usr/man(-/N) /(opt|usr)/(dt|share|X11R6|local)/(cat|)man(-/N) )
 
 # `sman' is the SGML manual directory for Solaris 7.
 
-if [[ $words[2] = (<->*|ln) ]]; then
+if [[ $words[2] = (<->*|l|n) ]]; then
   rep=(
   $manpath/(sman|man|cat)${words[2]}/${~approx}$PREFIX${~star}$SUFFIX.<->*(N:t) )
 else

Re: PATCH: _man copes with : delimited manpath

[Peter Stephenson]

Adam Spiers wrote:
> On my system, `manpath' generates colon-delimited paths, rather than
> space-delimited ones.  This patch solves that problem.  It also fixes
> what I think is a typo, but someone should check this.

You'll need to add this.  mp doesn't get split into words the first time
unless you use an array.

--- Completion/User/_man~	Mon Sep 27 15:41:48 1999
+++ Completion/User/_man	Mon Sep 27 15:42:53 1999
@@ -15,7 +15,7 @@
 fi
 
 if (( ! $#manpath )); then
-  local mp=$(manpath 2>/dev/null)
+  local mp=($(manpath 2>/dev/null))
   [[ "$mp" == *:* ]] && mp=( ${(s.:.)mp} )
   manpath=( $mp )
 fi

-- 
Peter Stephenson <pws@ibmth.df.unipi.it>       Tel: +39 050 844536
WWW:  http://www.ifh.de/~pws/
Dipartimento di Fisica, Via Buonarroti 2, 56127 Pisa, Italy

Re: PATCH: _man copes with : delimited manpath

[Tanaka Akira]

In article <9909271344.AA19699@ibmth.df.unipi.it>,
  Peter Stephenson <pws@ibmth.df.unipi.it> writes:

> +  local mp=($(manpath 2>/dev/null))

I think this is desired syntax. But it is not implemented yet.

is27e1u11% man <TAB>
_man: unknown file attribute [18]

Index: Completion/User/_man
===================================================================
RCS file: /projects/zsh/zsh/Completion/User/_man,v
retrieving revision 1.1.1.16
diff -u -F^( -r1.1.1.16 _man
--- Completion/User/_man	1999/09/27 14:24:58	1.1.1.16
+++ Completion/User/_man	1999/09/28 09:34:44
@@ -15,7 +15,8 @@
 fi
 
 if (( ! $#manpath )); then
-  local mp=($(manpath 2>/dev/null))
+  local mp
+  mp=($(manpath 2>/dev/null))
   [[ "$mp" == *:* ]] && mp=( ${(s.:.)mp} )
   manpath=( $mp )
 fi
-- 
Tanaka Akira

PATCH: Re: _man copes with : delimited manpath

[Oliver Kiddle]

Adam Spiers wrote:
> On my system, `manpath' generates colon-delimited paths, rather than
> space-delimited ones.  This patch solves that problem.  It also fixes
> what I think is a typo, but someone should check this.

Sorry, I wasn't thinking properly when I wrote it - all manpath commands I can find return a colon delimited string.

I prefered the more succinct way it was done so I'd suggest the following patch. An alternative would be to assign to MANPATH instead of manpath.

Oliver Kiddle

*** Completion/User/_man.bak	Wed Oct  6 17:09:43 1999
--- Completion/User/_man	Wed Oct  6 17:09:58 1999
***************
*** 14,25 ****
    approx="(#a${_comp_correct})"
  fi
  
! (( $#manpath )) || manpath=$(manpath 2>/dev/null) ||
    manpath=( /usr/man(-/N) /(opt|usr)/(dt|share|X11R6|local)/(cat|)man(-/N) )
  
  # `sman' is the SGML manual directory for Solaris 7.
  
! if [[ $words[2] = (<->*|ln) ]]; then
    rep=(
    $manpath/(sman|man|cat)${words[2]}/${~approx}$PREFIX${~star}$SUFFIX.<->*(N:t) )
  else
--- 14,25 ----
    approx="(#a${_comp_correct})"
  fi
  
! (( $#manpath )) || manpath=( ${(s.:.)$(manpath 2>/dev/null)} ) ||
    manpath=( /usr/man(-/N) /(opt|usr)/(dt|share|X11R6|local)/(cat|)man(-/N) )
  
  # `sman' is the SGML manual directory for Solaris 7.
  
! if [[ $words[2] = (<->*|l|n) ]]; then
    rep=(
    $manpath/(sman|man|cat)${words[2]}/${~approx}$PREFIX${~star}$SUFFIX.<->*(N:t) )
  else

Re: PATCH: Re: _man copes with : delimited manpath

[Adam Spiers]

Oliver Kiddle (opk@u.genie.co.uk) wrote:
> Adam Spiers wrote:
> > On my system, `manpath' generates colon-delimited paths, rather than
> > space-delimited ones.  This patch solves that problem.  It also fixes
                                                            ^^^^^^^^^^^^^
> > what I think is a typo, but someone should check this.
    ^^^^^^^^^^^^^^^^^^^^^^
[snip]

> ! if [[ $words[2] = (<->*|ln) ]]; then

[snip]

> ! if [[ $words[2] = (<->*|l|n) ]]; then

Was I wrong in thinking that this was a typo?  Or did you just
accidentally unfix my fix? :-)

Adam

P.S. Am I alone in finding context diffs remarkably awkward to read?

Re: PATCH: Re: _man copes with : delimited manpath

[Adam Spiers]

Adam Spiers, on thelonious (adam@thelonious.new.ox.ac.uk) wrote:
> Was I wrong in thinking that this was a typo?  Or did you just
> accidentally unfix my fix? :-)

Agh, ignore that; I read the patch the wrong way round.  Like I said ...

> P.S. Am I alone in finding context diffs remarkably awkward to read?

But it still looks this patch replaces two or more of the previous
_man patches.  Which ones?

Re: PATCH: Re: _man copes with : delimited manpath

[Oliver Kiddle]

Adam Spiers wrote:

> Agh, ignore that; I read the patch the wrong way round.  Like I said ...
> > P.S. Am I alone in finding context diffs remarkably awkward to read?

They're not too bad, the only problem is with a patch like mine where
there are only changed lines (as opposed to added/removed) so it is less
obvious which lines are new and which are old. You could always write a
quick script to colourise diffs to make it clearer. I use an awk script
as a filter before my pager when I use mush which colourises e-mails.

> But it still looks this patch replaces two or more of the previous
> _man patches.  Which ones?

It was intended to replace your original patch and as a result render
those which followed irrelevant. This is why I also included your typo
fix - so that people wouldn't have to apply half of your patch.

Oliver Kiddle