Function:     DCDPOS

Purpose:      The function dcdpos decodes position input for boxinp
              and posinp.

Category:     PHYSICAL COORDINATES

File:         cotrans.c

Author:       K.G. Begeman

Use:          INTEGER DCDPOS( SET    ,   input    character*(*)
                              SUBSET ,   input    integer
                              STRING ,   input    character*(*)
                              POS    ,   output   double precision array
                              OPTION )   input    integer

              DCDPOS      Depends on option. If OPTION greater than 0
                          (POSINP) DCDPOS returns the number of positions
                          decoded. If OPTION equals 0 (BOXINP) then DCDPOS
                          returns a coded number which can have the following
                          values:
                           0 - no position found in string
                           1 - one position found in string
                           2 - only a size found in string
                           3 - position and size found in string
                           4 - two positions found in string
                          In all case, a negative return value indicates
                          an error. The following errors might occur:
                          -1  - illegal use of 'PC', 'AC' or 'D'
                          -2  - prefix incompatible with axis
                          -3  - position incomplete
                          -4  - error reading descriptor info
                          -5  - 'D' not allowed for positions
                          -6  - No grid separation defined for units
                          -7  - Too many positions
                          -8  - Cannot obtain header information
                          -9  - No mixed epochs allowed
                          -10 - General decode error (detected by dcddble).
                          -11 - BLANKS decoded.
              SET         Input set name.
              SUBSET      Input subset coordinate word. Needed to
                          determine the dimension of the subset.
              STRING      String containing input to be decoded.
              POS         The decoded positions. The positions are stored
                          sequentially, meaning that the first N items
                          contain the first position (N is dimension of
                          subset), the second N items the second, etc.
              OPTION      If OPTION equals zero, then BOX decoding will
                          be done. If OPTION is greater than 0 then
                          position decoding will be done, where OPTION
                          indicated the maximum number of positions to
                          decode.

Updates:      Jan 18, 1990: KGB, Document created.
              Feb 17, 1992: KGB, Bug in trailing units repaired.
              Mar 27, 1995: KGB, Errors -10 and -11 implemented.