[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[netCDF #YDH-858665]: NetCDF variable name convention

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: [netCDF #YDH-858665]: NetCDF variable name convention
  • Date: Wed, 18 Jun 2008 06:29:07 -0600
Hi Paul,

> Is there a technical reason why the naming convention enforces a
> character (or underscore) as the first character or is this an
> arbitrary convention that could be relaxed?

Now that we have heard from users that this will cause a problem,
the restriction that names not begin with a number will be
removed in upcoming releases.

The technical reasons this change was made were to

 - match the Users Guide documentation that, since about 1990, has
   restricted names to begin with "_" or an alphabetic character

 - make the mapping from netCDF names to C and Fortran variable
   names in the generated code produced by "ncgen -c" and "ncgen -f"
   more natural (no C or Fortran names can begin with a number)

We didn't know anyone was beginning their names with a number, but
should have asked about this before making the change.  Also note
that the CF conventions require names to begin with an alphabetic
character, so beginning with a number means you data is technically
not  CF-compliant.  That may not be an issue, if using your data with
software that requires CF compliance is not required.

Thanks for the feedback.

--Russ

Russ Rew                                         UCAR Unidata Program
address@hidden                     http://www.unidata.ucar.edu



Ticket Details
===================
Ticket ID: YDH-858665
Department: Support netCDF
Priority: Normal
Status: Closed