[nclpy-group] Feedback on new feature...

[Rick Brownrigg]

Hi Alan,

This is an interesting and thoughtful suggestion. As an
instigator/implementer of the block-comment feature, I will certainly take
it under serious consideration.

Typically in language recognition applications, recognition of comments is
done at a very early stage of parsing - way before any notions of
"ncl-ness" are detected. Such is the case here. Providing feedback of a
block-comment in the interactive prompts may not be straightforward -- I'll
need to investigate further.

However, your comment invites another nice-to-have feature for interactive
usage: automatic indentation based upon nested levels of if/do constructs
(?)

Rick


On Tue, Jul 7, 2015 at 2:20 PM, Alan Brammer <abrammer at albany.edu> wrote:

> Coming back to this very briefly.
>
> I use ncl interactively a lot of the time while writing scripts.  If there
> was a way to indicate that one was within a blocked comment I think it
> would be useful.
>
> Currently I occasionally get stuck in nested do/if statements with no
> indication as to what level I'm at.  Block comments add another potential
> level of confusion.
>
> My initial thought would be to change the prompt when inside a block
> comment as below.
>
> ncl 1>
>
> to
>
> ncl 1;
>
>
>
>
> Alan.
>
>
>
>
> On Wed, Jul 1, 2015 at 6:53 PM, Maria Gehne - NOAA Affiliate <
> maria.gehne at noaa.gov> wrote:
>
>> Yes, the more forgiving to forgetfulness the better.
>>
>> Maria
>>
>> On Wed, Jul 1, 2015 at 4:42 PM, Dennis Shea <shea at ucar.edu> wrote:
>>
>>> Je suis d'accord ... D
>>>
>>> On Wed, Jul 1, 2015 at 3:44 PM, David Brown <dbrown at ucar.edu> wrote:
>>>
>>>> On Wed, Jul 1, 2015 at 3:13 PM, Alan Brammer <abrammer at albany.edu>
>>>> wrote:
>>>> > Hi all,
>>>> >
>>>> > My vote would have to be on the /;    ;/  notation for consistency
>>>> with
>>>> > existing comments.
>>>> >
>>>> > This would also hopefully mean you can leave the trailing ;/ when you
>>>> > uncomment a block. I know I would end up with a ton of trailing */
>>>> which
>>>> > would then trip errors otherwise.
>>>> >
>>>> That's an interesting observation, and in my mind adds points in favor
>>>> of the /; ;/ syntax.
>>>>  -dave
>>>>
>>>> >
>>>> > Alan.
>>>> >
>>>> >
>>>> >
>>>> > On Wed, Jul 1, 2015 at 2:42 PM, Karin Meier-Fleischer
>>>> > <meier-fleischer at dkrz.de> wrote:
>>>> >>
>>>> >> Hi to all,
>>>> >>
>>>> >> thanks to Mary for the welcome note, too. :-)
>>>> >>
>>>> >> I would prefer /* .. */ as used for C/C++.
>>>> >>
>>>> >> The /; .. ;/ looks a little bit strange and in my opinion the /* ..
>>>> */ is
>>>> >> much better to identify in scripts.
>>>> >> But that’s just my view.
>>>> >>
>>>> >> Bye,
>>>> >> Karin
>>>> >>
>>>> >>
>>>> >>
>>>> >> Am 01.07.2015 um 17:28 schrieb Rick Brownrigg <brownrig at ucar.edu>:
>>>> >>
>>>> >> Greetings everyone,
>>>> >>
>>>> >> Thanks to Mary for the welcoming introductory note.  We do have a
>>>> minor
>>>> >> feature in the works for NCL as a language, but one which I think
>>>> many will
>>>> >> find convenient:  block-comments, whereby one can readily comment out
>>>> >> multiple lines at once.
>>>> >>
>>>> >> The question to the group is:  what should that look like?
>>>> >>
>>>> >> As a C/C++/Java programmer, my natural inclination would be /*  ...
>>>> */
>>>> >> style bracketing comments.  However, I personally think that
>>>> notation looks
>>>> >> out of place in NCL code.  Given NCL's Fortran influences, we could
>>>> turn to
>>>> >> that language -- I'm surprised however that modern Fortran does not
>>>> have a
>>>> >> block-comment (?)
>>>> >>
>>>> >> With ";" being NCL's line-comment character, one suggestion that has
>>>> stuck
>>>> >> is
>>>> >>
>>>> >>    /;   ...  ;/  bracketing notation.
>>>> >>
>>>> >> To those of you who might not be C/C++/Java programmers, is there
>>>> another
>>>> >> more natural style appropriate for a language like NCL?
>>>> >>
>>>> >> We appreciate any comments...
>>>> >>
>>>> >> Rick
>>>> >>
>>>> >>
>>>> >>
>>>> >>
>>>> >>
>>>> >>
>>>> >> _______________________________________________
>>>> >> nclpy-group mailing list
>>>> >> nclpy-group at ucar.edu
>>>> >> List instructions, subscriber options, unsubscribe:
>>>> >> http://mailman.ucar.edu/mailman/listinfo/nclpy-group
>>>> >>
>>>> >
>>>> >
>>>> > _______________________________________________
>>>> > nclpy-group mailing list
>>>> > nclpy-group at ucar.edu
>>>> > List instructions, subscriber options, unsubscribe:
>>>> > http://mailman.ucar.edu/mailman/listinfo/nclpy-group
>>>> >
>>>> _______________________________________________
>>>> nclpy-group mailing list
>>>> nclpy-group at ucar.edu
>>>> List instructions, subscriber options, unsubscribe:
>>>> http://mailman.ucar.edu/mailman/listinfo/nclpy-group
>>>>
>>>
>>>
>>> _______________________________________________
>>> nclpy-group mailing list
>>> nclpy-group at ucar.edu
>>> List instructions, subscriber options, unsubscribe:
>>> http://mailman.ucar.edu/mailman/listinfo/nclpy-group
>>>
>>>
>>
>> _______________________________________________
>> nclpy-group mailing list
>> nclpy-group at ucar.edu
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/nclpy-group
>>
>>
>
> _______________________________________________
> nclpy-group mailing list
> nclpy-group at ucar.edu
> List instructions, subscriber options, unsubscribe:
> http://mailman.ucar.edu/mailman/listinfo/nclpy-group
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/nclpy-group/attachments/20150708/15adfc30/attachment.html