hi all,
I have some comments regarding 'any', having myself some experience in
both using and implementing it:
I agree with the idea of handling open type values like unions of TTCN,
but would like to question the necessity of introducing the TTCN type
'anytype', into the language. I would propose to omit the facility of
defining open types out of TTCN, and use the TTCN notation for
referencing types for accessing the type. Having imported a type 'T', I
would reference the type of open type field 'o' as follows:
T.o
Referenced like this, not as an 'anytype', the type is correct because,
e.g.
the ASN.1 open type may be restricted, and may be different than the
union of all known types.
To me the deciding question for including the 'anytype' in the language
is if there is any use for it for the purpose of testing, to:
- check that a value is of 'anytype',
- send a value of 'anytype',
- store a value of 'anytype',
- reference a value of 'anytype'.
As I see only the last action as meaningful, I would reason that the
existence of this type is the result of the way the data specifications
are written and that the type itself is not necessary in TTCN. Having
already a way to reference those types, I don't see why they should be
introduced into the TTCN type system.
Union access seems appropriate for handling values of open types,
although the BNF (revision 2002) seems to delete the operator
'IsChosen'. Satisfied with these facilities, I would define the
referenced open types in TTCN only by saying 'open or any types are used
as union types of TTCN, with the name of the selected type in place of
the union selector identifier'.
regards,
Martti Söderlund
--
This email address is being protected from spambots. You need JavaScript enabled to view it.
Nokia Research Center
ext Jens Grabowski wrote:
>
> Hi together,
>
> the ANYTYPE has been requested in order
> to be IDL compatible.
>
> Find enclosed a proposal for the ANYTYPE.
>
> Comments are wellcome.
>
> Regards
> Jens Grabowski
>
>
> Name: any-type.doc
> Type: WINWORD File (application/msword)
> any-type.doc Encoding: base64
> Description: any-type.doc
> Download Status: Not downloaded with message
--
Martti Söderlund
This email address is being protected from spambots. You need JavaScript enabled to view it.
Nokia Research Center / Mobile Networks Laboratory