A thought on implementation of literal functions, if they are accepted
P-1788,
David Gay has a suite of routines in C for conversion of
IEEE-754 floating point numbers into literal format, with
correct rounding (to nearest, to nearest less than or equal,
to nearest greater than or equal), and for conversion of
literals to IEEE-754 format (with rounding to nearest,
to nearest less than or equal, to nearest greater than or equal).
I tried my hand at using these some years ago, but I don't think
I got them to function correctly. (It may have been a matter
of the environment, the compiler, or options on the compiler.)
I recall the routine for floating point to literal I tried
was "gdtoa". It can be found in the package "fp" in Netlib.
(In the Netlib listing, there is also the following link:
http://www.cs.berkeley.edu/~jhauser/arithmetic/softfloat.html
to a software implementation of the IEEE floating point
standard.)
These items might be relevant to a reference implementation
of interval literal conversions.
Does anyone have experience with these?
Best regards,
Baker
--
---------------------------------------------------------------
Ralph Baker Kearfott, rbk@xxxxxxxxxxxxx (337) 482-5346 (fax)
(337) 482-5270 (work) (337) 993-1827 (home)
URL: http://interval.louisiana.edu/kearfott.html
Department of Mathematics, University of Louisiana at Lafayette
(Room 217 Maxim D. Doucet Hall, 1403 Johnston Street)
Box 4-1010, Lafayette, LA 70504-1010, USA
---------------------------------------------------------------