So I have these two functions: function task() { task_wrapper.pl $funcstack[1] "$@" } function tc() { task_wrapper.pl $funcstack[1] "$@“ } They are both wrappers for the same perl script which does its thing based on the value of $funcstack[1]; Works, but I’m wondering if I can spare myself the job remembering to create a new function for each and every new perl function I want to write. I could write a little script to pull out the subroutine names from the perl script and dump them into a file sourced by zsh and then reload zsh. Wondering if there might be some cool feature of zsh I don’t know about to assist with creating these functions on-the-fly.
On Tue, 2022-03-22 at 16:52 -0400, Steve Dondley wrote:
> So I have these two functions:
>
> function task() {
> task_wrapper.pl $funcstack[1] "$@"
> }
>
> function tc() {
> task_wrapper.pl $funcstack[1] "$@“
>
> }
>
> They are both wrappers for the same perl script which does its thing
> based on the value of $funcstack[1];
>
> Works, but I’m wondering if I can spare myself the job remembering to
> create a new function for each and every new perl function I want to
> write.
>
> I could write a little script to pull out the subroutine names from
> the perl script and dump them into a file sourced by zsh and then reload
> zsh.
>
> Wondering if there might be some cool feature of zsh I don’t know
> about to assist with creating these functions on-the-fly.
Recent versions of the shell allow you to copy functions from an old
name to a new name. It's efficiently implemented internally.
% oldname() { print I am $0: }
% oldname
I am oldname
% functions -c oldname newname
% newname
I am newname
pws
On 3/22/22, Steve Dondley <s@dondley.com> wrote:
> So I have these two functions:
>
> function task() {
> task_wrapper.pl $funcstack[1] "$@"
> }
>
> function tc() {
> task_wrapper.pl $funcstack[1] "$@“
>
> }
>
> They are both wrappers for the same perl script which does its thing based
> on the value of $funcstack[1];
>
> Works, but I’m wondering if I can spare myself the job remembering to create
> a new function for each and every new perl function I want to write.
>
> I could write a little script to pull out the subroutine names from the perl
> script and dump them into a file sourced by zsh and then reload zsh.
>
> Wondering if there might be some cool feature of zsh I don’t know about to
> assist with creating these functions on-the-fly.
You can define multiple functions like this:
function f1 f2 f3 f4() {
this body is shared by functions f1 f2 f3 and f4
}
and you can copy them like this (somewhat newer feature)
functions -c f1 g1
so if you can get all the function names in an array, you could just say:
task_func_names=( $(script to pull names out) )
function $task_func_names() {
task_wrapper.pl $funcstack[1] "$@"
}
--
Mikael Magnusson
Nice. This looks promising. Thanks!
> On Mar 22, 2022, at 5:03 PM, Mikael Magnusson <mikachu@gmail.com> wrote:
>
> On 3/22/22, Steve Dondley <s@dondley.com> wrote:
>> So I have these two functions:
>>
>> function task() {
>> task_wrapper.pl $funcstack[1] "$@"
>> }
>>
>> function tc() {
>> task_wrapper.pl $funcstack[1] "$@“
>>
>> }
>>
>> They are both wrappers for the same perl script which does its thing based
>> on the value of $funcstack[1];
>>
>> Works, but I’m wondering if I can spare myself the job remembering to create
>> a new function for each and every new perl function I want to write.
>>
>> I could write a little script to pull out the subroutine names from the perl
>> script and dump them into a file sourced by zsh and then reload zsh.
>>
>> Wondering if there might be some cool feature of zsh I don’t know about to
>> assist with creating these functions on-the-fly.
>
> You can define multiple functions like this:
> function f1 f2 f3 f4() {
> this body is shared by functions f1 f2 f3 and f4
> }
>
> and you can copy them like this (somewhat newer feature)
> functions -c f1 g1
>
> so if you can get all the function names in an array, you could just say:
> task_func_names=( $(script to pull names out) )
> function $task_func_names() {
> task_wrapper.pl $funcstack[1] "$@"
> }
>
> --
> Mikael Magnusson
>
On Tue, Mar 22, 2022 at 2:06 PM Mikael Magnusson <mikachu@gmail.com> wrote:
>
> so if you can get all the function names in an array, you could just say:
> task_func_names=( $(script to pull names out) )
> function $task_func_names() {
> task_wrapper.pl $funcstack[1] "$@"
> }
When using the "function" keyword, you don't need the empty parens,
either. That keyword further ignores the state of the MULTI_FUNC_DEF
option, which is what you want if you're using this array method.
OK, this is semi-working. The corresponding perl function is definitely getting called. If I it has a syntax error, STDERR gets output to the screen.
But for whatever reason, I can’t get any output on STDOUT with something as simple as:
sub blah {
system('echo', 'hi');
}
> On Mar 22, 2022, at 5:03 PM, Mikael Magnusson <mikachu@gmail.com> wrote:
>
> On 3/22/22, Steve Dondley <s@dondley.com> wrote:
>> So I have these two functions:
>>
>> function task() {
>> task_wrapper.pl $funcstack[1] "$@"
>> }
>>
>> function tc() {
>> task_wrapper.pl $funcstack[1] "$@“
>>
>> }
>>
>> They are both wrappers for the same perl script which does its thing based
>> on the value of $funcstack[1];
>>
>> Works, but I’m wondering if I can spare myself the job remembering to create
>> a new function for each and every new perl function I want to write.
>>
>> I could write a little script to pull out the subroutine names from the perl
>> script and dump them into a file sourced by zsh and then reload zsh.
>>
>> Wondering if there might be some cool feature of zsh I don’t know about to
>> assist with creating these functions on-the-fly.
>
> You can define multiple functions like this:
> function f1 f2 f3 f4() {
> this body is shared by functions f1 f2 f3 and f4
> }
>
> and you can copy them like this (somewhat newer feature)
> functions -c f1 g1
>
> so if you can get all the function names in an array, you could just say:
> task_func_names=( $(script to pull names out) )
> function $task_func_names() {
> task_wrapper.pl $funcstack[1] "$@"
> }
>
> --
> Mikael Magnusson
>
And here’s what I get with `type`: > $ type tbbbc tbbbc is a shell function from /User/me/.local/zsh/aliases/taskwarrior So the zsh function generated is definitely recognized as a function. > On Mar 22, 2022, at 5:09 PM, Steve Dondley <s@dondley.com> wrote: > > Nice. This looks promising. Thanks! > >> On Mar 22, 2022, at 5:03 PM, Mikael Magnusson <mikachu@gmail.com> wrote: >> >> On 3/22/22, Steve Dondley <s@dondley.com> wrote: >>> So I have these two functions: >>> >>> function task() { >>> task_wrapper.pl $funcstack[1] "$@" >>> } >>> >>> function tc() { >>> task_wrapper.pl $funcstack[1] "$@“ >>> >>> } >>> >>> They are both wrappers for the same perl script which does its thing based >>> on the value of $funcstack[1]; >>> >>> Works, but I’m wondering if I can spare myself the job remembering to create >>> a new function for each and every new perl function I want to write. >>> >>> I could write a little script to pull out the subroutine names from the perl >>> script and dump them into a file sourced by zsh and then reload zsh. >>> >>> Wondering if there might be some cool feature of zsh I don’t know about to >>> assist with creating these functions on-the-fly. >> >> You can define multiple functions like this: >> function f1 f2 f3 f4() { >> this body is shared by functions f1 f2 f3 and f4 >> } >> >> and you can copy them like this (somewhat newer feature) >> functions -c f1 g1 >> >> so if you can get all the function names in an array, you could just say: >> task_func_names=( $(script to pull names out) ) >> function $task_func_names() { >> task_wrapper.pl $funcstack[1] "$@" >> } >> >> -- >> Mikael Magnusson >> > >
OK, sorry for the noise. I introduced a bug in the script which prevented any functions from getting called. Everything is good. Thanks again!
> On Mar 22, 2022, at 7:28 PM, Steve Dondley <s@dondley.com> wrote:
>
> OK, this is semi-working. The corresponding perl function is definitely getting called. If I it has a syntax error, STDERR gets output to the screen.
>
> But for whatever reason, I can’t get any output on STDOUT with something as simple as:
>
> sub blah {
> system('echo', 'hi');
> }
>
>
>> On Mar 22, 2022, at 5:03 PM, Mikael Magnusson <mikachu@gmail.com> wrote:
>>
>> On 3/22/22, Steve Dondley <s@dondley.com> wrote:
>>> So I have these two functions:
>>>
>>> function task() {
>>> task_wrapper.pl $funcstack[1] "$@"
>>> }
>>>
>>> function tc() {
>>> task_wrapper.pl $funcstack[1] "$@“
>>>
>>> }
>>>
>>> They are both wrappers for the same perl script which does its thing based
>>> on the value of $funcstack[1];
>>>
>>> Works, but I’m wondering if I can spare myself the job remembering to create
>>> a new function for each and every new perl function I want to write.
>>>
>>> I could write a little script to pull out the subroutine names from the perl
>>> script and dump them into a file sourced by zsh and then reload zsh.
>>>
>>> Wondering if there might be some cool feature of zsh I don’t know about to
>>> assist with creating these functions on-the-fly.
>>
>> You can define multiple functions like this:
>> function f1 f2 f3 f4() {
>> this body is shared by functions f1 f2 f3 and f4
>> }
>>
>> and you can copy them like this (somewhat newer feature)
>> functions -c f1 g1
>>
>> so if you can get all the function names in an array, you could just say:
>> task_func_names=( $(script to pull names out) )
>> function $task_func_names() {
>> task_wrapper.pl $funcstack[1] "$@"
>> }
>>
>> --
>> Mikael Magnusson
>>
>
>