next up previous contents
Next: The Data Routine Up: PSF -- Point Previous: Introduction

How Does it Work?

The psf system chooses a routine to initialise a particular psf by looking at the name supplied at the PSF prompt by the user. If this name is a valid abbreviation of any known psf then the initialisation routine is called. The mapping between the full name and the routine is defined using the internal routine PSF0_DEFPSF in the PSF_INIT module. The initialisation routrine has the standard signature,

   SUBROUTINE psf_init_rtn( INTEGER psid, INTEGER status )
where psid is the psf identifier. The dataset with which the psf is associated can be extracted from the FileID member of the structure referenced by psid (see the file $AST_ETC/psf.adi for the structure definition).

The psf initialisation routine's job is to ask for or locate any additional information required to specify the psf when actual data is requested. This extra information is known as psf instance data as it is peculiar to the particular psf/dataset combination for which it is computed. This data can be stored in psid using the psf0_setid0t routines for subsequent retrieval by the routines evaluating the psf data.

The most important job in initialisation is to associate other routines to perform other functions in the psf system -- only one of these is mandatory, the data routine.



Asterix
Tue Oct 7 12:01:03 BST 1997