[J3] Question about SPLIT and TOKENIZE

[Bleikamp, Richard]

[AMD Official Use Only - General]

Happy New Year to everyone!

I believe Brad is correct, the behavior for when SET or STRING is zero length is implied (correctly),  but not explicitly stated (the entire contents of STRING is by definition a single token, including when STRING is zero length).

Malcolm, when can we start submitting edits for F202y?  Is the next meeting ok?  And do you agree a clarifying edit is appropriate?

Rich

From: J3 <j3-bounces at mailman.j3-fortran.org> On Behalf Of Brad Richardson via J3
Sent: Thursday, December 22, 2022 9:32 AM
To: General J3 interest list <j3 at mailman.j3-fortran.org>
Cc: Brad Richardson <everythingfunctional at protonmail.com>
Subject: Re: [J3] Question about SPLIT and TOKENIZE

Caution: This message originated from an External Source. Use proper caution when opening attachments, clicking links, or responding.

I will note that SCAN and VERIFY both have explicit behavior defined for zero length STRING and SET. It would probably be prudent to do the same here. If SET is zero length, it would make sense to treat the entirety of STRING as a single token. If STRING is zero length, it makes sense to consider it as a single, zero length token. I believe the current wording implies this, but it doesn't hurt to spell it out.

Regards,
Brad

On Thu, 2022-12-22 at 08:53 +0900, Malcolm Cohen via J3 wrote:
Hi folks,

It has been pointed out to me by Van Snyder that these don't make much sense when the SET argument is a zero-length string.

As I read the wording, it looks like these will just select the whole of STRING as a single token. So maybe it's not actually broken.

Should we require SET to have length greater than zero? (This would be a minor technical change if so.)

Cheers,
--
..............Malcolm Cohen, NAG Oxford/Tokyo.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.j3-fortran.org/pipermail/j3/attachments/20230102/ad4b97e1/attachment.htm>