Consider just the fragment
&I SETA 0
&J SETA 1
&F SETC ' '
.*
.MACDAT ANOP ,
&STR SETC '&&MDAT('.'&J'.')&F.SETA '
AIF (&I EQ 0).MACD30
&STR SETC '&STR'.'&MDAT(&I),'
.MACD30 ANOP ,
&STR SETC '&STR'.'&MDAT(&I+1),&MDAT(&I+2),'
&STR SETC '&STR'.'&MDAT(&I+3),&MDAT(&I+4),'
&STR SETC '&STR'.'&MDAT(&I+5),&MDAT(&I+6),'
&STR SETC '&STR'.'&MDAT(&I+7)'
. . .
It would appear that a sequence of strings of the form
|&whatever(1) seta 1,2,3,4,5,6,7,8
|&whatever(9) seta 9,10,11,12,13,14,15,16
with the values 1, 2, 3, . . . replaced by more interesting ones is being
formatted.
This can be accomplished by
|&nrows seta . . . --an approprtiate value
|&i seta 0 --initialize .oloop index
|.oloop anop , --top of outer, row loop
|&i seta &i+1 --increment .oloop index
|&rows_complete setb (&i gt &nrows) --last row string formatted?
| aif (&rows_complete).olend --if so, leave .oloop
|&string setc '&identifier.('.'&i'.') seta '
|&delimiter setc ',' --list-element delimiter, comma
|&rowbase seta (&i-1)*8 --0, 8, 16, . . .
|&j seta 0 --initialize .iloop index
|.iloop anop --top of inner, element loop
|&j seta &j+1 --increment .iloop index
|&eor setb (&j gt 8) --end-of-row reached?
| aif (&eor).ilend --if so, leave .iloop
|¬_8 setb (&j lt 8) --not last concatenand in row?
| aif (¬_8).delimiter_set --if so, ok
|&delim setc '' --no, nul out list-element delimiter
|.delimiter_set anop
|&esub seta &rowbase+&j --calculate row-element subscript
|&string setc '&string'.'&element(&esub)'.'&delim' --concatenate an element
| ago .iloop --next concatenard
|.ilend anop , --bottom of .iloop
| ago .oloop --next row da capo
|.olend anop , --bottom of .oloop
E. B. White wrote that anyone who admired the word 'personalized' should be
free to use it himself, but that he should not be free to teach others to do so.
The macro language is a powerful statement-level one that should be written
with respect for the facilities it makes available; in particular, it should be
written to be maintainable and readable. DO-group surrogates should be
explicit, subscript calculations should be explicit, etc., etc.
John Gilmore
