0008153: Extend usage of break and continue statements

(0016441)
Jens Grabowski (manager)
20-12-2022 12:49

Additional idea: Add label to loop header.

TTF discussion: To be discussed in the scope of the next TTF. Proposal cannot be implemented as proposed for select case statement. It would lead to a backwards incompatible change.

(0016542)
Jens Grabowski (manager)
07-11-2023 17:01

TTF discussion: Break and Continue shall be analysed in the scope of the major release. Additional ideas include the deprecation of Goto as well as introduction of breaks/continue across several scope units.

(0016570)
Gusztáv Adamis (developer)
09-11-2023 21:15

After diiscussions within TTF:

Possible solution can be "named cycles"
while (...) {
   while (...){
      break w1; //or continue w1;
      }
   } : w1

break w1; will terminate both cycles, execution continues with the next instruction after w1.
continue w1; will terminate the innermost cycle and takes the next iteration of the "w1" cycle.

Similar construct can be applied for nested alt instructions with break/repeat "named alt").

alt {
  [] ...{
          alt {
            [] ... {break a1;}
            ...
          }
        ...
  [] ....
        }
}: a1;

"altlabels" and "cyclelabels" can be mixed, until they do not "jump" out of the scope.

alt {
[] ... { while (...) { break a2;}}
...
}: a2

If this construct is implemented it may cause to deprecate the goto/label.

(0016581)
Gusztáv Adamis (developer)
10-11-2023 12:03

The resolution of the ticket shall be postponed to the next major revision.

(0016608)
Olivier Genoud (administrator)
26-01-2024 16:27

It seems to be another flavour of "goto", so we do not see the benefit. On the other hand, it is increasing the complexity of the language allowing further flavours of code writing. Also, we are not aware that code cannot be written using existing means.

Issue History
Date Modified	Username	Field	Change
11-12-2022 19:21	Matthias Simon	New Issue
11-12-2022 20:07	Matthias Simon	Summary	Add optional label to break and continue => Extend usage of break and continue statements
11-12-2022 20:07	Matthias Simon	Description Updated	View Revisions
11-12-2022 20:07	Matthias Simon	File Added: CR8153.docx
20-12-2022 12:49	Jens Grabowski	Note Added: 0016441
20-12-2022 12:49	Jens Grabowski	Assigned To	=> Jens Grabowski
20-12-2022 12:49	Jens Grabowski	Status	new => assigned
07-11-2023 17:01	Jens Grabowski	Note Added: 0016542
09-11-2023 21:15	Gusztáv Adamis	Note Added: 0016570
10-11-2023 12:03	Gusztáv Adamis	Note Added: 0016581
26-01-2024 16:27	Olivier Genoud	Note Added: 0016608

Notes
(0016441) Jens Grabowski (manager) 20-12-2022 12:49	Additional idea: Add label to loop header. TTF discussion: To be discussed in the scope of the next TTF. Proposal cannot be implemented as proposed for select case statement. It would lead to a backwards incompatible change.

(0016542) Jens Grabowski (manager) 07-11-2023 17:01	TTF discussion: Break and Continue shall be analysed in the scope of the major release. Additional ideas include the deprecation of Goto as well as introduction of breaks/continue across several scope units.

(0016570) Gusztáv Adamis (developer) 09-11-2023 21:15	After diiscussions within TTF: Possible solution can be "named cycles" while (...) { while (...){ break w1; //or continue w1; } } : w1 break w1; will terminate both cycles, execution continues with the next instruction after w1. continue w1; will terminate the innermost cycle and takes the next iteration of the "w1" cycle. Similar construct can be applied for nested alt instructions with break/repeat "named alt"). alt { [] ...{ alt { [] ... {break a1;} ... } ... [] .... } }: a1; "altlabels" and "cyclelabels" can be mixed, until they do not "jump" out of the scope. alt { [] ... { while (...) { break a2;}} ... }: a2 If this construct is implemented it may cause to deprecate the goto/label.

(0016581) Gusztáv Adamis (developer) 10-11-2023 12:03	The resolution of the ticket shall be postponed to the next major revision.

(0016608) Olivier Genoud (administrator) 26-01-2024 16:27	It seems to be another flavour of "goto", so we do not see the benefit. On the other hand, it is increasing the complexity of the language allowing further flavours of code writing. Also, we are not aware that code cannot be written using existing means.

Relationships

ETSI's Bug Tracker