A dataset with an table block of 10 rows and four columns

# --> begin
dataset "table.ds"
<
  attribute "ATT_STR" string "hello" "a familiar greeting" ""
  attribute "PI" real 3.1415 "a familiar number" "rad"

  table "TABLE" 10 "a small table"
  <
    attribute "GOOD" bool true "is the SAS good?" ""
    column "RAWX" int32 "raw CCD X coordinate" "pixel"
    column(32) "Strings" string "a string column" ""
    column(10,20) "MiniImages" real32 "images" "c/s"
    column(0) "VariableLength" int16 "as many as you line" "erg/s"
  >
>
# --> end

The following syntax rules apply:

• Table specifications have the following syntax:

  table name rows label

  where

  • name is the name of the table (a string).

  • rows is an integer specifying the number of rows in the table.

  • label is a comment string.

• Column specifications have the following syntax:

  column(dim1, dim2, ...) name type label unit

  where

  • dim1, dim2, ... are integers demoting the dimensions of the column. The dimension specifier can be omitted for normal columns. Variable length columns are given dimension 0. String columns can only have one dimension. If more than one dimension is specified, the column will be treated as a one-dimensional column with dimensions equal to the product of the of the dimensions given. This is a limitation of FITS.

  • name is the name of the column (a string).

  • type is the type of the data stored in the column. It may be one of int8 int16 int32 real32 real64 string bool.

  • label is a comment string.

  • unit is a unit string.