File Formats and Data Structures
Each of the individual tasks (flux generation, event generator, tracking,
response,...) must read input from the previous step, process it
in memory and write the new results out (and possibly carry through
information from the input).
- File Format
- is the organization of the data on disk (or tape).
This is how data is exchanged between the steps. Primary goals
should be:
- Compactness
- Transportability
- Completeness
- Data Structure
- is the representation, or layout, of data
in memory. This can be individual variables, arrays, whole
common blocks ({\sc fortran}) or more complex layouts (eg. C structures).
This is what each program works on. These should be arranged for:
- Ease of use (including meaningful names)
- Intuitiveness
- Full expression of data relationships
- Usually the two are closely related in form.
- Packing (unpacking) information
to be written out should strip out (reconstruct) redundant information if
possible. For each standard file format and a corresponding data structure,
there should be a (1) output routine (2) input routine (3) dump routine.
- It would be advantageous to standardize these at each level.
But we must remain flexible enough to evolve as necessary to meet
individual needs.
hatcher@astro.indiana.edu
- updated 1996.01.11 -
[an error occurred while processing this directive]