Format: Structure output lists

• The standard Format option is used in the READ and WRITE functions.
• ⇒ how to define formats for (file-) MatrixExplorer and SORT
• To display and edit a file or an array the DLG function also uses the Format option.
• Not all Format variations make sense in these 4 functions, e.g. there cannot be a display with binary format in DLG.
• Numerics and strings can be freely mixed in one statement
• integer and real formats can be used for all numeric data (numeric values are internally stored as 64 bit floating point
• for interactive examples see Tutorial.hic(Format) and Tutorial.hic(PictureFormat)

• floating point formats

  (Fortran standard)
  • WRITE(Format=fmt) 1, 21.9, 314.15, -4.51, 123456789 ! sample output real numbers

  	Syntax (*)	fmt sample	1	21.9	314.15	-4.51	123456789
  I integer	[r]Iw[.m]	5i10	1	21	314	-4	123456789
  F real	[r]Fw.d	5F10.2	1.00	21.90	314.15	-4.51	123456789
  E exponent	[r]Ew.d[Ee]	5E10.2	0.10E+01	0.22E+02	0.31E+03	-0.45E+01	0.12E+09
  ES scientific	[r]ESw.d[Ee]	5ES10.2	1.00E+00	2.19E+01	3.14E+02	-4.51E+00	1.23E+08
  EN engineering	[r]ENw.d	5EN10.2	1.00E+00	21.90E+00	314.15E+00	-4.51E+00	123.46E+06
  EN exp ctrl	[r]ENw.d[Ee]	5EN10.2e1	1.00E+0	21.90E+0	314.15E+0	-4.51E+0	123.46E+6
  G all	[r]Gw.d[Ee]	5G10.2	1.0	22.	0.31E+03	-4.5	0.12E+09

  (*) w=width, d=digits, r=repeat count, m=minimum places, Ee=exponential places, [optional]

• integer formats

  (Fortran standard)
  • WRITE(Format=fmt) 0, 1, 16, 65, 255 ! sample output integer numbers

  type	Syntax	fmt sample	0	1	16	65	255
  I integer	[r]Iw[.m]	5I10	0	1	16	65	255
  L logical	[r]Lw	5L10	F	T	T	T	T
  Z hexadecimal	[r]Zw[.m]	5Z10	0	1	10	41	FF
  O octal	[r]Ow[.m]	5O10	0	1	20	101	377
  B binary	[r]Bw.m	5B10.1	0	1	10000	1000001	11111111

• character format

  (Fortran standard)
  • WRITE(fmt="a, "(format embedded string)", 3x, a12") "automatic length", "(length limited)"

  automatic length (format embedded string)   (length limi

• To facilitate the

  formatting of column headers

  etc., HicEst allows to

  right-adjust strings

  by the use of numeric formats. Only the width is used then, e.g.

  WRITE(fmt="(I5, F8.2/)") "int", "value", 1, 2, 3, 4 ! this will output: 
    int   value 
      1    2.00
      3    4.00

• format control specifiers

  :
  • "," commas separate the list items as in the above examples
  • "/" slash instead comma starts next record
  • "nX" skips n positions
  • "[n](...)" to repeat the enclosed formats [n times]

• width directed output format

  :

  Format="Fw"

  (Fortran standard requires Fw.d) for output with

  maximum significance for given width

  :

  WRITE(Format="(F4)") 1, "abc", 1/2, "defg", 1/3, "hijklmn", 1/4 
  1    abc0.5 defg0.33hijk0.25

• free numeric input format

  :

  Format="F"

  , read the

  next available numeric values

  to scalar or arrays:
  • READ(Text="ABC=-1.23, Value2 is 4E3", Format="3F") a, b, c ! assigns a = -1.23, b=2, c=4000
  • READ(Text=string(1:30), items=n) vec ! n = elements found, the array vec is filled from left to right

• output format width as needed

  :

  Format="F"

  writes numeric values, maximum is 10 digits:
  • WRITE(Format="(F)") 1, 1/2, 1/3, 1/4
  • 1 0.5 0.3333333333 0.25
  or for a vector of dimension 12:
  • REAL :: vec(12)
  • vec = $^3 ! fills vec with the index cubes
  • WRITE(Fmt="(F)") vec ! will result in "1 8 27 64 125 216 343 512 729 1E3 1331 1728"

• binary formats

  : these internal formats work with character strings as well as with disk files. Binary formats can be mixed with all other formats.

  type	Format	range	decimal places
  signed binary 1-Byte integer	[r]ib1	-128 .. 127	2....3
  unsigned binary 1-Byte integer	[r]ibu1	0 .. 255	2....3
  signed binary 2-Byte integer	[r]ib2	-32768 .. 32767	4....5
  unsigned binary 2-Byte integer	[r]ibu2	0 .. 65535	4....5
  signed binary 4-Byte integer	[r]ib4	-2147483648.. 2147483647	9...10
  unsigned binary 4-Byte integer	[r]ibu4	0 .. 42949672959	9...10
  signed binary 8-Byte integer	[r]ib8	-9.22337E18.. 9.22337E18	18...19
  signed binary 4-Byte real	[r]fb4	±1.18E-38 .. ±3.40E38	6....7
  signed binary 4-Byte real	[r]fb4	±1.18E-38 .. ±3.40E38	6....7
  signed binary 8-Byte real	[r]fb8	±1.00E-300 .. ±1.00E300	15..16

  • WRITE(Text=tx, fmt="IB1,IB2") 65, 66+256*67 ! writes "ABC"
  • (CHAR(65)=="A", CHAR(66)=="B", CHAR(67)=="C")
  • READ(Text=tx, fmt="IB1,IB2") x, y ! makes x=65, y=17218 if tx was written with above statement

• time format

  : Format="U...":
  • read or write days in Microsoft (Excel) compatible days. Day 1 is Dec. 31, 1899
  • fractional days are transferred to hours, minutes, seconds, milliseconds
  • the strings following the "U" are case sensitive:

  • CC	YY	MM	DD	HH	mm	SS	sss	W or WW or WWW
    Century	Year	Month	Day	Hour	minute	Second	msec	day of the week

    • WRITE(Format="UDD.MM.CCYY" ) 1 ! days > 0: writes "31.12.1899"
    • WRITE(Format="UYY/MM/DD HH:mm" ) 0 ! days <=0: current day+time "07/09/22 19:06"
    • WRITE(Format="UHH:mm:SS sss msec") 0 ! writes "19:06:59 249msec"
    • WRITE(Format="UWWW MM-DD HH.mm") 0 ! writes: "Sun 12-22 19.06"
    • READ(text="03:00 1899/12/30", fmt="UHH:mm CCYY MM DD") day ! day --> 0.125
    • READ(text="03:00", fmt="UHH:mm") day ! default = 1899/12/30, day --> 0.125
    • READ(text="22.09.07 06:00", fmt="UDD.MM.YY HH:mm") day ! default century CC is 20, day --> 39347.25
    • READ(text="1899/12/29", f="UCCYY MM DD") day ! day --> -1
  • any other text can be freely interspersed.

• picture output format

  matches the output list to the given format string, like in:
  • WRITE(R=1, fmt=template) 79, "hello", pi, -678, "RESULT"

  123 Rightalign  1.2345 keep this  1.2E+01 C_this_is_the_template 
   79      hello  3.1416 keep this -6.8E+02        RESULT

  • template text tokens starting with L, C, or R is Left or Center or Right aligned
  • for an interactive demo see Tutorial.hic(PictureFormat)

How your donation will be used, and how to get a HicEst key code in return ©2000-2010 Georg Petrich, HicEst Instant Prototype Computing. All rights reserved. Impressum