[fitsbits] BINTABLE convention for >999 columns

[Mark Taylor]

Tom,

just one point of detail on your comments:

On Mon, 10 Jul 2017, Tom McGlynn (NASA/GSFC Code 660.1) wrote:

> 2. A  solution compatible with the current FITS standard that gives a couple
> of orders of magnitude increase in range:  Using base 36 indices in some
> format.
> -- This is pretty backwards compatible, but there can be problems with any
> FITS file that might have used keywords that happens to have the same first 5
> characters as some column  keyword.  If would actually make any existing FITS
> table that had, e.g., the keyword TFORMATS illegal ( Since for TFORMnnn must
> have nnn < TFIELDS).  This will require rather a lot of changes to code since

No, that's not how I'd envisage it.  In the convention I'm suggesting,
XXXXXaaa (with aaa in [A-Z]) keywords are only significant as
column metadata in a table which flags its conformance by appropriate
use of the XT_ICOL and XT_NCOL columns.  Unless the XT_ICOL/XT_NCOL
keywords are in place, processing proceeds goes on as before.
So you can use TFORMATS to mean whatever you like in a normal BINTABLE.

It's *possible* that some existing FITS BINTABLE has XT_ICOL and XT_NCOL
that look like what I'm suggesting in this convention, but I think
very unlikely.

Actually, I envisaged it so that only those columns within the
XT_NCOL range would be affected, so that TFORMATS would be
available for unrestricted use in any wide table with fewer than
1511 columns.  I *thought* that was in accordance with normal
BINTABLE usage, but on closer reading of the standard I see that
TFORMnnn columns are forbidden for nnn>TFIELDS, so probably
this convention should follow the same model.

Mark

--
Mark Taylor   Astronomical Programmer   Physics, Bristol University, UK
m.b.taylor at bris.ac.uk +44-117-9288776  http://www.star.bris.ac.uk/~mbt/