Why [tcl::mathfunc::min 1 2] but [expr min(1,2)] with commas?

[Busirane]

I use min and max infrequently enough that, when I do, I always trip over the differing syntax of the expr and mathfunc versions. Why does [expr min()] want commas? Especially since it *allows* whitespace, as long as there's a comma too. It's almost like [expr min()] expects a single string argument, but that's not how it's documented, at least in 8.5.13.

[Gerald Lester]

On 09/22/2017 09:14 AM, Busirane wrote:
> I use min and max infrequently enough that, when I do, I always trip over the differing syntax of the expr and mathfunc versions. Why does [expr min()] want commas? Especially since it *allows* whitespace, as long as there's a comma too. It's almost like [expr min()] expects a single string argument, but that's not how it's documented, at least in 8.5.13.
>

Because [expr] has its own mini-language (as do regular expressions) and
the [expr[ mini-language is an infix language and not a prefix one.

In short, it is what it is for historical reasons.

--
+----------------------------------------------------------------------+
| Gerald W. Lester, President, KNG Consulting LLC |
| Email: Gerald...@kng-consulting.net |
+----------------------------------------------------------------------+

[Georgios Petasis]

Στις 22/9/2017 18:35, ο Gerald Lester έγραψε:
> On 09/22/2017 09:14 AM, Busirane wrote:
>> I use min and max infrequently enough that, when I do, I always trip
>> over the differing syntax of the expr and mathfunc versions. Why does
>> [expr min()] want commas? Especially since it *allows* whitespace, as
>> long as there's a comma too. It's almost like [expr min()] expects a
>> single string argument, but that's not how it's documented, at least
>> in 8.5.13.
>>
>
> Because [expr] has its own mini-language (as do regular expressions) and
> the [expr[ mini-language is an infix language and not a prefix one.
>
> In short, it is what it is for historical reasons.
>

Also, the functions in tcl::mathfunc::* are the underline implementation
of expr's functions. Adding new functions in the tcl::mathfunc
namespace, will make these functions available through expr.

For example:

proc tcl::mathfunc::test_func {a} {puts $a; incr a}
expr test_func(8)

8
9

George

[Rich]

Likely because traditional mathematical functional notation uses commas
to separate function arguments. So 'expr' functional notation uses
commas at the separator to maintain a semblance of familarity in using
the functions.

The Tcl proc's in the mathfunc namespace are Tcl proc's, so they play
by Tcl proc argument rules, not mathematical functional notation rules.

[Busirane]

On Friday, September 22, 2017 at 11:35:42 AM UTC-4, Gerald Lester wrote:
> it is what it is for historical reasons.

That's actually the answer I expected when I asked the question. Thanks, Gerald.