On Mon, Apr 2, 2018 at 12:38 PM, Erich Steinböck <erich.steinbo...@gmail.com
> wrote:
> I've added preliminary "static" addess-with tests (they don't execute real
> commands)
> I also changed READY to READY: because this is what open returns and what
> the STREAM options tests for.
>
> The tests show a few quirks (marked with @@ in the test group), like:
>
> - should `address '' '' with` be allowed with no actual redirection?
>
> That's probably not what you're thinking it is. That would be considered
the implicit ADDRESS VALUE form, but yes, WITH should be supported and no
type should raise an error. I suspect I missed adding the WITH check to
that parth.
>
> - REPLACE never seems to work for OUTPUT, e. g. `address "" "" with
> output replace stem stem. `, not only for USING STREAM where it is
> explicitly forbidden
>
> Still working through unit tests...will look into.
>
> - have the FIFO / LIFO redirections been intentionally left out?
>
> That must be a Regina extension...those are not part of the ANSI
standard. I will take a look at what Regina does and will add if I think
it makes sense.
Rick
>
> -
>
>
> On Mon, Apr 2, 2018 at 3:01 PM, Rick McGuire <object.r...@gmail.com>
> wrote:
>
>> Gil,
>>
>> The addressWith sandbox version now builds cleanly. Basically, everything
>> is in place except for the code in SystemCommands.cpp that performs the I/O
>> redirects with the shell. As part of this, I added a dummy TEST address
>> handler to SystemCommands so I can start writing unit tests to test of the
>> interpreter side of this code. This dummy version also shows how to use the
>> new APIs I've added.
>>
>> A few comments/questions:
>>
>> 1) I don't know if you need this or not, but I added an API that will let
>> you know if OUTPUT and ERROR are redirected to the same target. If that's
>> the case, the output will be interleaved if to used both Write apis to add
>> data, but I don't know if that knowledge helps with setting up with the
>> piping or not. This was trivial to add, so it was no big deal.
>>
>> 2) Should I consider adding RexxQueue to the list of types supported by
>> USING? This would be a fairly simple addition.
>>
>> 3) Right now, ADDRESS WITH will only work with command handlers that use
>> the new APIs I just created. Currently, if the target address handler is
>> not one that supports I/O redirection, I just ignore the specification and
>> go ahead and issue the command. Should this be considered an error? It
>> would be fairly trivial make this an error condition.
>>
>> Rick
>>
>> ------------------------------------------------------------
>> ------------------
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>> _______________________________________________
>> Oorexx-devel mailing list
>> Oorexx-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/oorexx-devel
>>
>>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> Oorexx-devel mailing list
> Oorexx-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/oorexx-devel
>
>
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Oorexx-devel mailing list
Oorexx-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/oorexx-devel
