Hi All,
I am just catching up on the backlog of CF emails. My sense too is that this discussion is trying to solve the problems caused by a lack of grammar with alternatives and/or stopgaps. My preference is to overcome the grammar/vocab challenge, but I am well aware that an accepted solution has not yet occurred.
In order to get us on the right track, I propose we take advantage of Jonathan's suggestion in a way that doesn't require a full grammar/vocab definition, and doesn't require any changes to the controlling CF documents.
Specifically, I propose the following:
1) We leverage Jonathan's grammar program into (a) a program that checks a proposed std_name by parsing it to see whether it fits existing grammar/vocab rules, and/or (b) a std_name generation program.
2) Std_names are still proposed in the ordinary way, but if they have passed the checker or been created through the generator then it will be easy for people to accept them. We might even move to a mode in which pre-approved std_names are automatically accepted after a month, unless someone objects.
This has several advantages:
A) It will reduce time and effort by everyone to get std_names approved.
B) Neither the parser nor the generator needs to be complete (ie, it is OK if some existing names don't comply, or there are some valid new cases they don't cover)
C) Proposals that don't fit the standard construction can still be approved, and will highlight ways to complete and extend the parser/generator.
D) Any mistakes by the parser/generator should be caught by the email list.
I see no disadvantages other than the need for someone to create the parser and/or generator, which should be technically straightforward.
Best wishes,
Philip
-----------------------------------------------------------------------
Dr Philip Cameron-Smith, pjc at llnl.gov, Lawrence Livermore National Lab.
-----------------------------------------------------------------------
> -----Original Message-----
> From: CF-metadata [mailto:cf-metadata-bounces at cgd.ucar.edu] On Behalf Of
> Jonathan Gregory
> Sent: Monday, September 17, 2012 9:20 AM
> To: cf-metadata at cgd.ucar.edu
> Subject: Re: [CF-metadata] Another potentially useful extension to the
> standard_name table
>
> Dear Martin, Roy, John, Robert
>
> Reading the last few days' emails all at once I have may have skipped important
> details; if so, apologies for that. I too am in favour of a grammar, such as my
> earlier attempt http://climate.ncas.ac.uk/~jonathan/CF_metadata/14.1/
> Robert subsequently coded this grammar in an appropriate software language.
> This grammar has only one level of patterns, but some of its lexicon could be
> reduced further by having more than one level.
>
> A grammar would be useful for constructing standard names. People proposing
> names could be offered menus that allowed them to suggest names that
> followed existing patterns, or extensions to vocabulary, or new patterns.
> Thus all standard_names would naturally exist both as a specification that
> consists of a pattern with specific vocabulary items filling certain place- holders
> (the semantic tags, in effect), and as a joined-up standard_name. It's equivalent
> information. The specification could also be automatically translated into the
> accompanying description, since each pattern or semantic tag could trigger an
> appropriate descriptive text.
>
> It's easier to construct standard_names than to parse them, although parsing is
> possible. Hence it may be useful to give software access to the spec as well as
> the joined-up name.
>
> Best wishes
>
> Jonathan
> _______________________________________________
> CF-metadata mailing list
> CF-metadata at cgd.ucar.edu
> http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
Received on Fri Sep 21 2012 - 16:52:33 BST
