compinit as function?

compinit as function?

[Andrej Borsenkow]

Is there any real reason, why compinit cannot be a function?

The only problem as far as I can see is, that it defines several global
AA's - and current typeset's will make them local.

One workaround is to use

: ${(AA)hash:=}

(it is ugly, but it works)

Another possibility is to add the global flag to typeset: typeset -g. Can it
be useful in other cases?

Having compinit as function will allow to simply drop the whole stuff in
auoloaded fpath - and don't worry about path to compinit anymore. Of course,
then some other standard path to dump file is needed.

/andrej

Re: compinit as function?

[Peter Stephenson]

"Andrej Borsenkow" wrote:
> Is there any real reason, why compinit cannot be a function?

One reason is that it wastes memory, since it's quite large and you only
use it once.

> Another possibility is to add the global flag to typeset: typeset -g. Can it
> be useful in other cases?

I've been thinking about this.  It should be easy to add.


$testhash is just what it's name suggests.  It was added because there's
currently no other way of checking that special associative arrays work
properly (in particular being restored after `testhash=... cmd').  It will
eventually disappear.

-- 
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: compinit as function?

[Andrej Borsenkow]

> "Andrej Borsenkow" wrote:
> > Is there any real reason, why compinit cannot be a function?
>
> One reason is that it wastes memory, since it's quite large and you only
> use it once.
>

It is not as large. The main part are defines for compdef and compconf :-)
And may be more options to compinit can be useful in future ... compinit
rehash to reread new fpath as example? If Zsh will ever be able to execute
functions on access to variables ...

/andrej