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

Re: CIF specification: reserved prefixes

On Wednesday 29 October 2003 17:55, Brian McMahon wrote:
>
> > Probably misunderstandings on my part, but I had the impression that
> > _data_name_tags had no semantic meaning of their own and parsing them is
> > therefore against CIF philosophy. All the meaning is embedded in the
> > associated dictionary definition for which there should be a precise
> > string match.
>
> Strictly, that's correct. Reserved prefixes simply guard against name
> collisions by encouraging people to create tags with their own unique
> prefix. I suppose my thoughts are running along the lines of specialist
> applications (that maybe I'll get round to writing one day) that help
> the registry to screen out candidate prefixes that would collide with
> existing "official" category names - i.e. it would not be helpful if
> someone tried to register "atom" as a private prefix.
Hi Brian,

So  prefixes like atom_site and atom_type are objectionable because
they have underscores and coincidently are category names,
but it would be perfectly ok to use refl.hkl as a prefix (registered or 
otherwise) because it has no underscore?
If  "atom" was used as a prefix you can't exclude it on iether grounds.
It is neither a category nor contains an underscore, but is merely a leading 
component of many recognisable dictionary defined tags. 
I suspect also that if E. Dowty ever decided to export his ATOMS archives as 
CIFs he would probably rather appreciate using "atoms" as a prefix. That 
would not give a match against the dictionary but may be equally unhelpful to
casual readers of the CIF.

For the purposes of parsing CIF tags, it looks like banning underscores from 
prefixes gains you nothing that you couldn't get more reliably and 
efficiently by introducing a standard prefix separator such as the  ":" 
character.


best wishes
Doug

_______________________________________________
comcifs mailing list
comcifs@iucr.org
http://scripts.iucr.org/mailman/listinfo/comcifs

Reply to: [list | sender only]