Re: why not promote the plague ? (brace|rc|alternative|modern) syntax

Let's try this again:

First, we have the standard syntax that works in any POSIX shell:
>
>     for var in val1 val2 val3; do command1; command2; done
>     while command1; command2; do command3; command4; done
>     until command1; command2; do command3; command4; done
>     if command1; command2; then command3; command4; else command5;
> command6; fi
>
> The C-style *for* loop is not part of POSIX, but is common to bash, ksh,
> and zsh:
>
>     for (( init; condition; increment )) do command1; command2; done
>
> The *repeat* loop is purely a Zsh innovation not present in the other
> shells:
>
>     repeat $times; do command1; command2; done
>

>
All of the above have alternative syntax in Zsh that works irrespective of
> the *SHORT_LOOPS* option.
>

So far, so correct, but I got a little too specific with the details.

> Since all of these constructs have the form *intro-keyword *part1 *keyword1
> *part2 *keyword2*, the basic idea is to replace *keyword1 *and *keyword2 *with
> curly braces.  But the *for*...*in *loop also has a *keyword0*, namely
> *in*, which needs its own d
>

>     for var (val1 val2 val3) { command1; command2; }
>     while (command1; command2) { command3; command4; }
>     until (command1; command2) { command3; command4; }
>     if (command1; command2) { command3; command4; } else { command5;
> command6; }
>     for (( init; condition; increment )) { command1; command2; }
>     repeat $times { command1; command2; }
>
> NOTE: If the conditional command for while/until/if is a single ((...)) or
> [[...]] test, you don't need the extra parentheses around it.
>
> Finally, we have the Zsh "short" syntax, which works only with *setopt
> SHORT_LOOPS*, and *only if the body is a single command*:
>
>     for var (val1 val2 val3) command1;
>     while (command1; command2) command3
>     until (command1; command2) command3
>     if (command1; command2) command3;
>     for (( init; condition; increment )) command1;
>
> NOTE: again, the parentheses can be omitted around a single ((...)) or
> [[...]] test. Also, as far as I know, there's no way to attach an *else* to
> an *if* when using the short syntax.
>
> On Sat, Dec 14, 2024 at 2:29 AM Marc Chantreux <[email protected]> wrote:
>
>> On Fri, Dec 13, 2024 at 07:07:17PM -0500, Lawrence Velázquez wrote:
>> > Mikael's point was to distinguish between the alternative forms
>> > and the SHORT_LOOPS forms.  None of your examples works without
>> > SHORT_LOOPS
>>
>> thanks for noticing. which leads me to open zshmisc(1) and read
>> read "ALTERNATE FORMS FOR COMPLEX COMMANDS". TIL that my coding
>> style is actually a mix of the alternative form and SHORT_LOOPS
>> and the difference is subtile:
>>
>>         « The  short  versions  below only work if sublist is of the form
>> `{
>>         list }' or if the SHORT_LOOPS option is set »
>>
>> regards
>> --
>> Marc Chantreux
>>
>>
>
> --
> Mark J. Reed <[email protected]>
>


-- 
Mark J. Reed <[email protected]>