Re: [TCLCORE] [Ext] Re: TIP 672 Implementation Complete - Ready for Sponsorship

[Florent M. <flo...@gm...>]

Another alternative proposal :

set x [($y+1)]

It is very readable, easy to write, there is eventually a unicode version

set x 〖$y+1〗

1 . We define a new kind of substitution

RULE [7'] Expression substitution :

If a word contains an open bracket *«** [ »* immediately followed by an 
open parenthesis « ( », then Tcl performs an Expression substitution. To 
do this it invokes the Tcl interpreter recursively to process the 
characters following the open parenthesis as a Tcl mathematical 
Expression. The Expression must be terminated by a close bracket « *] » 
*immediately preceded by a closed parenthesis « ) ». The result of the 
expression is substituted into the word in place of the 
brackets-parenthesis and all of the characters between them. There may 
be any number of expression substitutions in a single word. Expression 
substitution is not performed on words enclosed in braces.

2. We create the test in ParseToken

... else if (*src == '[') {

            if (src[1] == '(') {

                      // Expression substitution

                      // must finish by « )] »

                     // doesn't know how to do it...

             } else {

                      // Command substitution

             }

3. we create a -noexpr switch in the subst command

....

Examples :

inline expression :  string range "my long sentence" 0 end-[($x+2)]

With native list expr enancement :  .c create rectangle {*}[( $X , $Y, 
$X+100, $Y+100 )]

With multiples instructions enhancement :  set multiexpr [( x = 1; y = 
2; $x-$y, $y-$x )]

Readable, easy to type, test logically positionned in the C source code, 
no double substitution, avoiding confusion with variable substitution : 
This is the solution !

Florent


Le 22/10/2025 à 15:52, Zaumseil René a écrit :
>
> And how will this be documented? Imho it should be mentioned in the 
> dodecalogue.
>
> Currently the used syntax is only used as array variable and not as 
> math command.
>
> I remember the big discussion on {*} syntax here.
>
> Again, is this change really more readable and useable?
>
> Is the benefit worth the additional syntax.
>
> Why not create real benefit with p.e. a multiple set command ala tip 674.
>
> Or a new math command without need of $.. access to variables.
>
> Regards
>
> rene
>
> *Von:*Florent Merlet <flo...@gm...>
> *Gesendet:* Mittwoch, 22. Oktober 2025 15:39
> *An:* tc...@ro...
> *Cc:* et...@ro...; tcl...@li...
> *Betreff:* [Ext] Re: [TCLCORE] TIP 672 Implementation Complete - Ready 
> for Sponsorship
>
> So, everybody in the core team agree with the implementation of a new 
> math-Expression substitution as a memlink command into the procedure 
> in charge of variable name substitution ?
>
> Le mer. 22 oct. 2025, 14:07, EricT <tw...@gm...> a écrit :
>
>     Hi everyone,
>
>     I wanted to update you on the status of TIP 672 (expression
>     substitution syntax).
>
>     The prototype implementation is now complete and available at:
>     https://github.com/rocketship88/tcl-tip-672-prototype
>
>     The implementation has been parameterized to support both
>     discussed approaches with a simple compile-time flag:
>     - Mode 1: $(expr) syntax
>     - Mode 2: $=(expr) or $^(expr) - configurable character
>
>     Key features:
>     - Comprehensive test suite (82 tests covering edge cases)
>     - Minimal code changes (~100 lines)
>     - Clean integration with existing parser
>     - Full Tcl test suite passes (only 3 expected differences)
>
>     There is one known issue: a minor memory leak in the synthetic
>     command strings that will need attention during integration.
>
>     At this point, I need a sponsor to upload the code to a TIP branch
>     on Fossil so the core team can review and decide whether to adopt
>     the TIP and, if so, which syntax to use. I'm available for any
>     collaborative work, questions, or modifications needed.
>
>     I'm grateful for all the feedback and support from the community.
>     I look forward to the next steps in the process.
>
>     Best regards,
>
>     Eric
>
>     _______________________________________________
>     Tcl-Core mailing list
>     Tcl...@li...
>     https://lists.sourceforge.net/lists/listinfo/tcl-core
>