NAME
    Data::SecsPack - pack and unpack numbers in accordance with SEMI E5-94

SYNOPSIS
     #####
     # Subroutine interface
     #  
     use Data::SecsPack qw(bytes2int config float2binary 
                        ifloat2binary int2bytes   
                        pack_float pack_int pack_num  
                        str2float str2int 
                        unpack_float unpack_int unpack_num);

     $big_integer = bytes2int( @bytes );

     $old_value = config( $option );
     $old_value = config( $option => $new_value);

     ($binary_magnitude, $binary_exponent) = float2binary($magnitude, $exponent, @options); 
     ($binary_magnitude, $binary_exponent) = float2binary($magnitude, $exponent, [@options]); 
     ($binary_magnitude, $binary_exponent) = float2binary($magnitude, $exponent, {@options}); 
 
     ($binary_magnitude, $binary_exponent) = ifloat2binary($imagnitude, $iexponent, @options);
     ($binary_magnitude, $binary_exponent) = ifloat2binary($imagnitude, $iexponent, [@options]);
     ($binary_magnitude, $binary_exponent) = ifloat2binary($imagnitude, $iexponent, {@options});

     @bytes = int2bytes( $big_integer );

     ($format, $floats) = pack_float($format, @string_integers);

     ($format, $integers) = pack_int($format, @string_integers);

     ($format, $numbers, @string) = pack_num($format, @strings);

     $float = str2float($string);
     (\@strings, @floats) = str2float(@strings);

     $integer = str2int($string);
     (\@strings, @integers) = str2int(@strings);

     \@ingegers = unpack_int($format, $integer_string); 
     \@floats   = unpack_float($format, $float_string); 
     \@numbers  = unpack_num($format, $number_string); 

     #####
     # Class interface
     #
     use Data::SecsPack;

     $big_integer = bytes2int( @bytes );

     ($binary_magnitude, $binary_exponent) = float2binary($magnitude, $exponent, @options); 
     ($binary_magnitude, $binary_exponent) = float2binary($magnitude, $exponent, [@options]); 
     ($binary_magnitude, $binary_exponent) = float2binary($magnitude, $exponent, {@options}); 

     ($binary_magnitude, $binary_exponent) = ifloat2binary($imagnitude, $iexponent, @options);
     ($binary_magnitude, $binary_exponent) = ifloat2binary($imagnitude, $iexponent, [@options]);
     ($binary_magnitude, $binary_exponent) = ifloat2binary($imagnitude, $iexponent, {@options});

     @bytes = int2bytes( $big_integer );

     $secspack = new Data::Secs2( @options );
     $secspack = new Data::Secs2( [@options] );
     $secspack = new Data::Secs2( {options} );

     ($format, $floats) = Data::SecsPack->pack_float($format, @string_integers);

     ($format, $integers) = Data::SecsPack->pack_int($format, @string_integers);

     ($format, $numbers, @strings) = Data::SecsPack->pack_num($format, @strings);

     $integer = Data::SecsPack->str2int($string)
     (\@strings, @integers) = Data::SecsPack->str2int(@strings);

     $float = Data::SecsPack->str2float($string);
     (\@strings, @floats) = Data::SecsPack->str2float(@strings);

     \@ingegers = Data::SecsPack->unpack_int($format, $integer_string); 
     \@floats   = Data::SecsPack->unpack_float($format, $float_string); 
     \@numbers  = Data::SecsPack->unpack_num($format, $number_string); 

DESCRIPTION
    The subroutines in the "Data::SecsPack" module packs and unpacks numbers
    in accordance with SEMI E5-94. The E5-94 establishes the standard for
    communication between the equipment used to fabricate semiconductors and
    the host computer that controls the fabrication. The equipment in a
    semiconductor factory (fab) or any other fab contains every conceivable
    known microprocessor and operating system known to man. And there are a
    lot of specialize real-time embedded processors and speciallize
    real-time embedded operating systems in addition to the those in the PC
    world.

    The communcication between host and equipment used packed nested list
    data structures that include arrays of characters, integers and floats.
    The standard has been in place and widely used in china, germany, korea,
    japan, france, italy and the most remote places on this planent for
    decades. The basic data structure and packed data formats have not
    changed for decades.

    This stands in direct contradiction to common conceptions of many in the
    Perl community. The following quote is taken from page 761, *Programming
    Perl* third edition, discussing the "pack" subroutine:

    "Floating-point numbers are in the native machine format only. Because
    of the variety of floating format and lack of a standard "network"
    represenation, no facility for interchange has been made. This means
    that packed floating-point data written on one machine may not be
    readable on another. That is a problem even when both machines use IEEE
    floating-point arithmetic, because the endian-ness of memory
    representation is not part of the IEEE spec."

    SEMI E5-94 and their precessors do standardize the endian-ness of
    floating point, the packing of nested data, used in many programming
    languages, and much, much more. The nested data has many performance
    advantages over the common SQL culture of viewing and representing data.
    The automated fabs of the world make use of nested data not only for
    communication between machines but also for local processing at the host
    and equipment.

    The endian-ness of SEMI E5-94 is the first MSB byte. Maybe this is
    because it makes it easy to spot numbers in a packed data structure.

    Does this standard communications protocol ensure that everything goes
    smoothly without any glitches with this wild mixture of hardware and
    software talking to each other in real time? Of course not. Bytes get
    reverse. Data gets jumbled from point A to point B. Machine time is
    non-existance. Big ticket, multi-million dollar fab equipment has to
    work to earn its keep. And, then there is the everyday business of
    suiting up, with humblizing hair nets, going through air and other
    showers just to get in to the clean room. And make sure not to do
    anything that will damage a little cassette containing a million dollars
    worth of product. It is totally amazing that the product does get out
    the door.

  SECSII Format

    The Data::SecsPack suroutines packs and unpacks numbers in accordance
    with SEMI E5-94, Semiconductor Equipment Communications Standard 2
    (SECS-II), avaiable from

     Semiconductor Equipment and Materials International
     805 East Middlefield Road,
     Mountain View, CA 94043-4080 USA
     (415) 964-5111
     Easylink: 62819945
     http://www.semiconductor-intl.org
     http://www.reed-electronics.com/semiconductor/
 
    The format of SEMI E5-94 numbers are established by below Table 1.

                   Table 1 Item Format Codes

     unpacked   binary  octal  hex   description
     ---------------------------------------------------------
     T          001001   11    0x24  Boolean
     S8         011000   30    0x60  8-byte integer (signed)
     S1         011001   31    0x62  1-byte integer (signed)
     S2         011010   32    0x64  2-byte integer (signed)
     S4         011100   34    0x70  4-byte integer (signed)
     F8         100000   40    0x80  8-byte floating
     F4         100100   44    0x90  4-byte floating
     U8         101000   50    0xA0  8-byte integer (unsigned)
     U1         101001   51    0xA4  1-byte integer (unsigned)
     U2         101010   52    0xA8  2-byte integer (unsigned)
     U4         101100   54    0xB0  4-byte integer (unsigned)

    Table 1 complies to SEMI E5-94 Table 1, p.94, with an unpack text symbol
    and hex columns added. The hex column is the upper Most Significant Bits
    (MSB) 6 bits of the format code in the SEMI E5-94 item header (IH)

    In accordance with SEMI E5-94 6.2.2,

    1   the Most Significat Byte (MSB) of numbers for formats S2, S4, S8,
        U2, U4, U8 is sent first

    2   the signed bit for formats F4 and F8 are sent first.

    3   Signed integer formats S1, S2, S4, S8 are two's complement

    The memory layout for Data::SecsPack is the SEMI E5-94 "byte sent first"
    has the lowest memory address.

    The SEMI E5-94 F4 format complies to IEEE 754-1985 float and the F8
    format complies to IEEE 754-1985 double. The IEEE 754-1985 standard is
    available from:

     IEEE Service Center
     445 Hoe Lane,
     Piscataway, NJ 08854
  
    The SEMI E5-94 F4, IEEE 754-1985 float, is 32 bits with the bits
    assigned follows:

     S EEE EEEE EMMM MMMM MMMM MMMM MMMM MMMM

    where S = sign bit, E = 8 exponent bits M = 23 mantissa bits

    The format of the float S, E, and M are as follows:

    Sign of the number
        The sign is one bit, 0 for positive and 1 for negative.

    exponent
        The exponent is 8 bits and may be positive or negative. The IEEE 754
        exponent uses excess-127 format. The excess-127 format adds 127 to
        the exponent. The exponent is re-created by subtracting 127 from the
        exponent.

    Magnitude of the number
        The magnitude or mantissa is a 23 bit unsigned binary number where
        the radix is adjusted to make the magnitude fall between 1 and 2.
        The magnitude is stored ignoring the 1 and filling in the trailing
        bits until there are 23 of them.

    The SEMI E5-94 F4, IEEE 754-1985 double, is 64 bits with S,E,M as
    follows: S = sign bit, E = 11 exponent bits M = 52 mantissa bits

    The format of the float S, E, and M are as follows:

    Sign of the number
        The sign is one bit, 0 for positive and 1 for negative.

    exponent
        The exponent is 8 bits and may be positive or negative. The IEEE 754
        exponent uses excess-1027 format. The excess-1027 format adds 1027
        to the exponent. The exponent is re-created by subtracting 1027 from
        the exponent.

    Magnitude of the number
        The magnitude or mantissa is a 52 bit unsigned binary number where
        the radix is adjusted to make the magnitude fall between 1 and 2.
        The magnitude is stored ignoring the 1 and filling in the trailing
        bits until there are 52 of them.

    For example, to find the IEEE 754-1985 float of -10.5

    *   Convert -10.5 decimal to -1010.1 binary

    *   Move the radix so magitude is between 1 and 2, -1010. binary to
        -1.0101 * 2^ +3

    *   IEEE 754-1985 sign is 1

    *   The magnitude dropping the one and filling in with 23 bits is

         01010000000000000000000

    *   Add 127 to the exponent of 3 to get

         130 decimal converted to 8 bit binary 

         10000010

    *   Combining into IEEE 754-1985 format:

         11000001001010000000000000000000

         1100 0001 0010 1000 0000 0000 0000 0000

         C128 0000 hex

SUBROUTINES
  bytes2int

     $big_integer = bytes2int( @bytes );

    The "bytes2int" subroutine counvers a "@bytes" binary number with the
    Most Significant Byte (MSB) $byte[0] to a decimal string number
    "$big_integer" using the "Data::BigInt" program module. As such, the
    only limitations on the number of binary bytes and decimal digits is the
    resources of the computer.

  config

     $old_value = config( $option );
     $old_value = config( $option => $new_value);
     (@all_options) = config( );

    The "config" subroutine reads and writes the default, startup options
    for the subroutines in the "Data::Secs2" program module and package. The
    options, with description in the subroutine where they are used, are as
    follows:

     used by                                        
     subroutine    option                        default value
     ----------------------------------------------------------
                   big_float_version
                   big_int_version
                   version

     bytes2int 

     float2binary  decimal_integer_digits          20
                   extra_decimal_fraction_digits    5 

     ifloat2binary decimal_fraction_digits         25
                   binary_fraction_bytes           10

     int2bytes   
     pack_float 
     pack_int 

     pack_num       nomix                          0
 
     str2float
     str2int 
     unpack_float
     unpack_int
     unpack_num

    The "bin_float_version" "bin_int_version" and "version" configuration
    variables are the versions for the "Math::BigFloat" "Math::BigInt"
    "Data::Secs2" program modules respectively.

  float2binary

     ($binary_magnitude, $binary_exponent) = float2binary($magnitude, $exponent, @options); 

    The "ifloat2binary" subroutine converts a decimal float with a base ten
    "$magnitude" and "$exponent" to a binary float with a base two
    "$binary_magnitude" and "$binary_exponent".

    The "ifloat2binary" assumes that the decimal point is set by "ixpeonent"
    so that there is one decimal integer digit in "imagnitude" The
    "ifloat2binary" produces a "$binary_exponent" so that the first byte of
    "$binary_magnitude" is 1 and the rest of the bytes are a base 2
    fraction.

    The "float2binary" subroutine uses the "ifloat2binary" for the small
    $exponents part and the native float routines to correct the
    "ifloat2binary" for the base ten exponent factor outside the range of
    the "ifoat2binary" subroutine.

    The "float2binary" subroutine uses the options "decimal_integer_digits",
    "$decial_fraction_digits", "extra_decimal_fraction_digits" in
    determining the "$iexponent" passed to the "ifloat2binary" subroutine.
    The option "decimal_integer_digits" is the largest positive base ten
    "$iexponent" while smallest "$ixponent" is the half
    "$decial_fraction_digits" + "extra_decimal_fraction_digits". The
    "float2binary" subroutine "extra_decimal_fraction_digits" only for
    negative "$iexponent". The "float2binary" subroutine uses any base ten
    "$exponent" from "$iexponent" breakout to adjust the "ifloat2binary"
    subroutine results using native float arith.

  ifloat2binary

     ($binary_magnitude, $binary_exponent) = ifloat2binary($imagnitude, $iexponent, @options);

    The "$ifloat2binary" subroutine converts a decimal float with a base ten
    "$imagnitude" and "$iexponent" using the "Math::BigInt" program module
    to a binary float with a base two "$binary_magnitude" and a base two
    "$binary_exponent". The "$ifloat2binary" assumes that the decimal point
    is set by "ixpeonent" so that there is one decimal integer digit in
    "imagnitude" The "ifloat2binary" produces a "$binary_exponent" so that
    the first byte of "$binary_magnitude" is 1 and the rest of the bytes are
    a base 2 fraction.

    Since all the calculations use basic integer arith, there are practical
    limits on the computer resources. Basically the limit is that with a
    zero exponent, the decimal point is within the significant "imagnitude"
    digits. Within these limitations, the accuracy, by chosen large enough
    limits for the binary fraction, is perfect.

    The first step of the "ifloat2binary" subroutine is zero out "iexponent"
    by breaking up the "imagnitude" into an integer part "integer" and
    fractional part "fraction" consist with the "iexponent". The
    c<ifloat2binary> will add as many significant decimal zeros to the right
    of "integer" in order to zero out "iexponent"; likewise it will add as
    many decimal zeros to the left of "integer" to zero out "exponent"
    within the limit set by the option "decimal_fraction_digits". If
    "ifloat2binary" cannot zero out "iexponent" without violating the
    "decimal_fraction_digits", "ifloat2binary" will discontinue processing
    and return an "undef" "$binary_magnitude" with and error message in
    "$binary_exponent".

    This design is based on the fact that the conversion of integer decimal
    to binary decimal is one to one, while the conversion of fractional
    decimal to binary decimal is not. When converting from decimal fractions
    with finite digits to binary fractions repeating binary fractions of
    infinity size are possible, and do happen quite frequently. An unlimited
    repeating binary fraction will quickly use all computer resources. The
    "binary_fraction_bytes" option provides this ungraceful event by
    limiting the number of fractional binary bytes. The default limits of 20
    "decimal_fraction_digits" and "binary_fraction_bytes" 10 bytes provides
    a full range of 0 - 255 for each binary byte. The ten bytes are three
    more bytes then are ever used in the largest F8 SEMI float format.

    The the following example illustrates the method used by "ifloat2binary"
    to convert decimal fracional digits to binary fractional bytes. Convert
    a 6 digit decimal fraction string into a binary fraction as follows:

     N[0-999999]      
     -----------  =  
       10^6          

     byte0    byte1   byte2    256         R2
     ----- +  ----- + ----- + ----- * ------------
     256^1    256^2   256^3   256^4     10 ^ 6

    Six digits was chosen so that the integer arith, using a 256 base, does
    not over flow 32 bit signed integer arith

     256 *   99999     =   25599744
     256 *  999999     =  255999744
     signed 32 bit max = 2147483648 / 256 = 8377608
     256 * 9999999     = 2559999744

    Note with quad arith this technique would yield 16 decimal fractional
    digits as follows:

     256 * 9999999999999999  =  2559999999999999744
     signed 64 bit max       =  9223372036854775808 / 256 = 36028797018963868
     256 * 99999999999999999 = 25599999999999999744

     Thus, need to get quad arith running.

     Basic step

      1      256 * N[0-999999]     1                     R0[0-999744]
     --- *   ----------------  =  ---- ( byte0[0-255] + ------------ ) 
     256         10 ^ 6           256                     10^6

    The results will have a range of

      1
     ---- ( 0.000000 to 255.999744)
     256 

    The fractional part, R0 is a six-digit decimal. Repeating the basic step
    three types gives the desired results. QED.

     2nd Iteration

      1      256 * R0[0-999744]       1                   R1[0-934464]
     --- *   --------------      =  ---- ( byte1[0-255] + ------------) 
     256         10 ^ 6              256                    10^6

     3rd Iteration

      1      256 * R1[0-934464]       1                   R2[0-222784]
     --- *   --------------      =  ---- ( byte2[0-239] + ------------) 
     256         10 ^ 6              256                    10^6

    Taking this out to ten bytes the first six decimal digits N[0-999999]
    yields bytes in the following ranges:

     byte    power      range    10^6 remainder
     ------------------------------------------ 
       0     256^-1     0-255    [0-999744]
       1     256^-2     0-255    [0-934464]
       2     256^-3     0-239    [0-222784]
       3     256^-4     0-57     [0-032704]
       4     256^-5     0-8      [0-372224]
       5     256^-6     0-95     [0-293440]
       6     256^-7     0-75     [0-120640]
       7     256^-8     0-30     [0-883840]
       8     256^-9     0-226    [0-263040]
       9     256^-10    0-67     [0-338249]

    The first two binary fractional bytes have full range. The rest except
    for byte 9 are not very close. This makes one wonder about the accuracy
    loss in translating from binary fractions to decimal fractions. One
    wonders just why have all theses problems with not just binary and
    decimal factions but fractions in general. Isn't mathematics wonderful.

    For example in convert from decimal to binary fractions there is no
    clean one to one conversion as for integers. For example, look at the
    below table of conversions:

     -1    -2     -3     -4     -5     binary power as a decimal   
     0.5   0.25  0.125 0.0625 0.03125  decimal power 
                                       decimal 
      0     0      0      0      0     0.00000
      0     0      0      0      1     0.03125
      0     0      0      1      1     0.0625
      0     0      1      0      0     0.125
      0     0      1      0      1     0.15625
      0     0      1      1      0     0.1875
      0     0      1      1      1     0.21875
      1     0      0      0      0     0.50000

  int2bytes

     @bytes = int2bytes( $big_integer );

    The "int2bytes" subroutine uses the "Data:BigInt" program module to
    convert an integer text string "$bit_integer" into a byte array,
    "@bytes", the Most Significant Byte (MSB) being "$bytes[0]". There is no
    limits on the size of "$big_integer" or "@bytes" except for the
    resources of the computer.

  new

     $secspack = new Data::Secs2( @options );
     $secspack = new Data::Secs2( [@options] );
     $secspack = new Data::Secs2( {options} );

    The "new" subroutine provides a method to set local options once for any
    of the other subroutines. The options may be modified at any time by
    "$secspack-"config($option => $new_value)>. Calling any of the
    subroutines as a "$secspack" method will perform that subroutine with
    the options saved in "secspack".

  pack_float

     ($format, $floats) = pack_float($format, @string_integers);

    The "pack_int" subroutine takes an array of strings, <@string_integers>,
    and a float format code, as specifed in the above "Item Format Code
    Table", and packs all the integers, decimals and floats as a float the
    "$format" in accordance with "SEMI E5-94". The "pack_int" subroutine
    also accepts the format code "F" and format codes with out the
    bytes-per-element number and packs the numbers in the format using the
    less space. In any case, the "pack_int" subroutine returns the correct
    "$format" of the packed "$integers".

    When the "pack_float" encounters an error, it returns "undef" for
    "$format" and a description of the error as "$floats".

  pack_int

     ($format, $integers) = pack_int($format, @string_integers);

    The "pack_int" subroutine takes an array of strings, <@string_integers>,
    and a format code, as specifed in the above "Item Format Code Table" and
    packs the integers, "$integers" in the "$format" in accordance with
    "SEMI E5-94". The "pack_int" subroutine also accepts the format code "I
    I1 I2 I8" and format codes with out the bytes-per-element number and
    packs the numbers in the format using the less space, with unsigned
    preferred over signed. In any case, the "pack_int" subroutine returns
    the correct "$format" of the packed "$integers".

    When the "pack_int" encounters an error, it returns "undef" for
    "$format" and a description of the error as "$integers". All the
    "@string_integers" must be valid Perl numbers.

  pack_num

     ($format, $numbers, @strings) = pack_num($format, @strings);

    The "pack_num" subroutine takes leading numbers in "@strings" and packs
    them in the "$format" in accordance with "SEMI E5-94". The "pack_num"
    subroutine returns the stripped "@strings" data naked of all leading
    numbers in "$format".

    The "pack_num" subroutine also accepts "$format" of "I I1 I2 I4 F" For
    these format codes, "pack_num" is extremely liberal and accepts
    processes all numbers consistence with the "$format" and packs one or
    more numbers in the "SEMI E5-94" format that takes the least space. In
    this case, the return $format is changed to the "SEMI E5-94" from the
    "Item FOrmat Code Table" of the packed numbers.

    For the "I" "$format", if the "nomix" option is set option, the
    "pack_num" subroutine will pack all leading, integers, decimals and
    floats as multicell float with the smallest space; otherwise, it will
    stop at the first decimal or float encountered and just pack the
    integers.

    The "pack_num" subroutine processes "@strings" in two steps. In the
    first step, the "pack_num" subroutine uses "str2int" and/or "str2float"
    subroutines to parse the leading numbers from the "@strings" as follows:

     ([@strings], @integers) = str2int(@strings); 
     ([@strings], @floats) = str2float(@strings); 

    In the second step, the "pack_num" subroutine uses "pack_int" and/or
    "pacK_float" to pack the parsed numbers.

  str2float

     $float = str2float($string);
     (\@strings, @floats) = str2float(@strings);

    The "str2float" subroutine, in an array context, supports converting
    multiple run of integers, decimals or floats in an array of strings
    "@strings" to an array of integers, decimals or floats, "@floats". It
    keeps converting the strings, starting with the first string in
    "@strings", continuing to the next and next until it fails an
    conversion. The "str2int" returns the stripped string data, naked of all
    integers, in "@strings" and the array of integers "@integers".

    In a scalar context, it parse out any type of $number in the leading
    "$string". This is especially useful for "$string" that is certain to
    have a single number.

  str2int

     $integer = str2int($string);
     (\@strings, @integers) = str2int(@strings); 

    In a scalar context, the "Data::SecsPack" program module translates an
    scalar string to a scalar integer. Perl itself has a documented
    function, '0+$x', that converts a scalar to so that its internal storage
    is an integer (See p.351, 3rd Edition of Programming Perl). If it cannot
    perform the conversion, it leaves the integer 0. Surprising not all
    Perls, some Microsoft Perls in particular, may leave the internal
    storage as a scalar string.

    The scalar "str2int" subroutine is basically the same except if it
    cannot perform the conversion to an integer, it returns an "undef"
    instead of a 0. Also, if the string is a decimal or floating point, it
    will return an undef. This makes it not only useful for forcing an
    integer conversion but also for testing a scalar to see if it is in fact
    an integer scalar. The scalar "str2int" is the same and supercedes
    C&<Data::SecsPack::str2int>. The "Data::SecsPack" program module
    superceds the "Data::SecsPack" program module.

    The "str2int" subroutine, in an array context, supports converting
    multiple run of integers in an array of strings "@strings" to an array
    of integers, "@integers". It keeps converting the strings, starting with
    the first string in "@strings", continuing to the next and next until it
    fails a conversion. The "str2int" returns the remaining string data in
    "@strings" and the array of integers "@integers".

  unpack_float

     \@floats   = unpack_float($format, $float_string);

    The "unpack_num" subroutine unpacks an array of floats "$float_string"
    packed in accordance with SEMI-E5 "$format". A valid "$format", in
    accordance with the above "Item Format Code Table", is "F4 F8".

    The "unpack_num" returns a reference, "\@floats", to the unpacked float
    array or scalar error message "$error". To determine a valid return or
    an error, check that "ref" of the return exists or is '"ARRAY"'.

  unpack_int

     \@integers = unpack_int($format, $integer_string); 

    The "unpack_num" subroutine unpacks an array of numbers
    "$string_numbers" packed in accordance with SEMI-E5 "$format". A valid
    "$format", in accordance with the above "Item Format Code Table", is "S1
    S2 S4 U1 U2 U4 T".

    The "unpack_num" returns a reference, "\@integers", to the unpacked
    integer array or scalar error message "$error". To determine a valid
    return or an error, check that "ref" of the return exists or is
    '"ARRAY"'.

  unpack_num

     \@numbers  = unpack_num($format, $number_string); 

    The "unpack_num" subroutine unpacks an array of numbers "$number_string"
    packed in accordance with SEMI-E5 "$format". A valid "$format", in
    accordance with the above "Item Format Code Table", is "S1 S2 S4 U1 U2
    U4 F4 F8 T". The "unpack_num" subroutine uses either "unpack_float" or
    "unpack_int" depending upon "$format".

    The "unpack_num" returns a reference, "\@numbers", to the unpacked
    number array or scalar error message "$error". To determine a valid
    return or an error, check that "ref" of the return exists or is
    '"ARRAY"'.

REQUIREMENTS
    Coming soon.

DEMONSTRATION
     #########
     # perl SecsPack.d
     ###

     ~~~~~~ Demonstration overview ~~~~~

    Perl code begins with the prompt

     =>

    The selected results from executing the Perl Code follow on the next
    lines. For example,

     => 2 + 2
     4

     ~~~~~~ The demonstration follows ~~~~~

     =>     use File::Package;
     =>     my $fp = 'File::Package';

     =>     my $uut = 'Data::SecsPack';
     =>     my $loaded;

     =>     my ($result,@result);

     =>     #########
     =>     # Subroutines to test that actual values are within
     =>     # and expected tolerance of the expected value
     =>     #
     =>     sub tolerance
     =>     {
     =>         my ($actual,$expected) = @_;
     =>         2 * ($expected - $actual) / ($expected + $actual);
     =>     }

     =>     sub pass_fail_tolerance
     =>     {   my ($actual,$expected) = @_;
     =>          (-$expected < $actual) && ($actual < $expected) ? 1 : 0;
     =>     }

     =>     my $tolerance_result;
     =>     my $float_tolerance = 1E-10;

     => ##################
     => # UUT Loaded
     => # 
     => ###

     =>    my $errors = $fp->load_package($uut, 
     =>        qw(bytes2int float2binary 
     =>           ifloat2binary int2bytes   
     =>           pack_float pack_int pack_num  
     =>           str2float str2int 
     =>           unpack_float unpack_int unpack_num) );
     => $errors
     ''

     => ##################
     => # str2int(\'033\')
     => # 
     => ###

     => $result = $uut->str2int('033')
     '27'

     => ##################
     => # str2int(\'0xFF\')
     => # 
     => ###

     => $result = $uut->str2int('0xFF')
     '255'

     => ##################
     => # str2int(\'0b1010\')
     => # 
     => ###

     => $result = $uut->str2int('0b1010')
     '10'

     => ##################
     => # str2int(\'255\')
     => # 
     => ###

     => $result = $uut->str2int('255')
     '255'

     => ##################
     => # str2int(\'hello\')
     => # 
     => ###

     => $result = $uut->str2int('hello')
     undef

     => ##################
     => # str2int(' 78 45 25', ' 512E4 1024 hello world') \@numbers
     => # 
     => ###

     => my ($strings, @numbers) = str2int(' 78 45 25', ' 512E4 1024 hello world')
     => [@numbers]
     [
               '78',
               '45',
               '25'
             ]

     => ##################
     => # str2int(' 78 45 25', ' 512E4 1024 hello world') \@strings
     => # 
     => ###

     => join( ' ', @$strings)
     '512E4 1024 hello world'

     => ##################
     => # str2float(' 78 -2.4E-6 0.25', ' 512E4 hello world') numbers
     => # 
     => ###

     => ($strings, @numbers) = str2float(' 78 -2.4E-6 0.0025', ' 512E4 hello world')
     => [@numbers]
     [
               [
                 '78',
                 '1'
               ],
               [
                 '-24',
                 '-6'
               ],
               [
                 '025',
                 -3
               ],
               [
                 '512',
                 '6'
               ]
             ]

     => ##################
     => # str2float(' 78 -2.4E-6 0.25', ' 512E4 hello world') \@strings
     => # 
     => ###

     => ($strings, @numbers) = str2float(' 78 -2.4E-6 0.0025', ' 512E4 hello world')
     => join( ' ', @$strings)
     'hello world'

     =>      my @test_strings = ('78 45 25', '512 1024 100000 hello world');
     =>      my $test_string_text = join ' ',@test_strings;
     =>      my $test_format = 'I';
     =>      my $expected_format = 'U4';
     =>      my $expected_numbers = '0000004e0000002d000000190000020000000400000186a0';
     =>      my $expected_strings = ['hello world'];
     =>      my $expected_unpack = [78, 45, 25, 512, 1024, 100000];

     =>      my ($format, $numbers, @strings) = pack_num('I',@test_strings);

     => ##################
     => # pack_num($test_format, $test_string_text) format
     => # 
     => ###

     => $format
     'U4'

     => ##################
     => # pack_num($test_format, $test_string_text) numbers
     => # 
     => ###

     => unpack('H*',$numbers)
     '0000004e0000002d000000190000020000000400000186a0'

     => ##################
     => # pack_num($test_format, $test_string_text) \@strings
     => # 
     => ###

     => [@strings]
     [
               'hello world'
             ]

     => ##################
     => # unpack_num($expected_format, $test_string_text) error check
     => # 
     => ###

     => ref(my $unpack_numbers = unpack_num($expected_format,$numbers))
     'ARRAY'

     => ##################
     => # unpack_num($expected_format, $test_string_text) numbers
     => # 
     => ###

     => $unpack_numbers
     [
               '78',
               '45',
               '25',
               '512',
               '1024',
               '100000'
             ]

     =>  
     =>      @test_strings = ('78 4.5 .25', '6.45E10 hello world');
     =>      $test_string_text = join ' ',@test_strings;
     =>      $test_format = 'I';
     =>      $expected_format = 'F8';
     =>      $expected_numbers = '405380000000000040120000000000003fd0000000000000422e08ffca000000';
     =>      $expected_strings = ['hello world'];
     =>      my @expected_unpack = (78, 4.5, 0.25,6.45E10);

     =>      ($format, $numbers, @strings) = pack_num('I',@test_strings);

     => ##################
     => # pack_num($test_format, $test_string_text) format
     => # 
     => ###

     => $format
     'F8'

     => ##################
     => # pack_num($test_format, $test_string_text) numbers
     => # 
     => ###

     => unpack('H*',$numbers)
     '405380000000000040120000000000003fd0000000000000422e08ffca000000'

     => ##################
     => # pack_num($test_format, $test_string_text) \@strings
     => # 
     => ###

     => [@strings]
     [
               'hello world'
             ]

     => ##################
     => # unpack_num($expected_format, $test_string_text) error check
     => # 
     => ###

     => ref($unpack_numbers = unpack_num($expected_format,$numbers))
     'ARRAY'

     => $unpack_numbers
     [
               '7800000000000017486e-17',
               '4500000000000006245e-18',
               '25e-2',
               '64500000000000376452e-9'
             ]

QUALITY ASSURANCE
  Test Report

      => perl SecsPack.t
 
     1..23
     # Running under perl version 5.006001 for MSWin32
     # Win32::BuildNumber 635
     # Current time local: Sat Apr 24 01:43:15 2004
     # Current time GMT:   Sat Apr 24 05:43:15 2004
     # Using Test.pm version 1.24
     # Test::Tech    : 1.2
     # Data::Secs2   : 1.17
     # Data::SecsPack: 0.03
     # =cut 
     ok 1 - UUT Loaded 
     ok 2 - str2int('033') 
     ok 3 - str2int('0xFF') 
     ok 4 - str2int('0b1010') 
     ok 5 - str2int('255') 
     ok 6 - str2int('hello') 
     ok 7 - str2int(' 78 45 25', ' 512E4 1024 hello world') @numbers 
     ok 8 - str2int(' 78 45 25', ' 512E4 1024 hello world') @strings 
     ok 9 - str2float(' 78 -2.4E-6 0.25', ' 512E4 hello world') numbers 
     ok 10 - str2float(' 78 -2.4E-6 0.25', ' 512E4 hello world') @strings 
     ok 11 - pack_num(I, 78 45 25 512 1024 100000 hello world) format 
     ok 12 - pack_num(I, 78 45 25 512 1024 100000 hello world) numbers 
     ok 13 - pack_num(I, 78 45 25 512 1024 100000 hello world) @strings 
     ok 14 - unpack_num(U4, 78 45 25 512 1024 100000 hello world) error check 
     ok 15 - unpack_num(U4, 78 45 25 512 1024 100000 hello world) numbers 
     ok 16 - pack_num(I, 78 4.5 .25 6.45E10 hello world) format 
     ok 17 - pack_num(I, 78 4.5 .25 6.45E10 hello world) numbers 
     ok 18 - pack_num(I, 78 4.5 .25 6.45E10 hello world) @strings 
     ok 19 - unpack_num(F8, 78 4.5 .25 6.45E10 hello world) error check 
     ok 20 - unpack_num(F8, 78 4.5 .25 6.45E10 hello world) float 0 
     ok 21 - unpack_num(F8, 78 4.5 .25 6.45E10 hello world) float 1 
     ok 22 - unpack_num(F8, 78 4.5 .25 6.45E10 hello world) float 2 
     ok 23 - unpack_num(F8, 78 4.5 .25 6.45E10 hello world) float 3 
     # Passed : 23/23 100%
 
  Other Tests

    The test script "SecsPackStress.t" provides a more thorough test and is
    provided in the distribution package along with its demo script
    companion "SecsPackStress.d".

    The Software Test Description (STD) for "SecsPackStress" is
    "SecsPackStress.pm" also in the distribution package. The installation
    runs both "SecsPack.t" and "SecsPackStress.t".

  Test Software

    The module "t::Data::SecsPack" is the Software Test Description(STD)
    module for the "Data::SecsPack". module.

    To generate all the test output files, run the generated test script,
    run the demonstration script and include it results in the
    "Data::SecsPack" POD, execute the following in any directory:

     tmake -test_verbose -replace -run  -pm=t::Data::SecsPack

    Note that tmake.pl must be in the execution path "$ENV{PATH}" and the
    "t" directory containing "t::Data::SecsPack" on the same level as the
    "lib" directory that contains the "Data::SecsPack" module.

NOTES
  AUTHOR

    The holder of the copyright and maintainer is

    <support@SoftwareDiamonds.com>

  COPYRIGHT NOTICE

    Copyrighted (c) 2002 Software Diamonds

    All Rights Reserved

  BINDING REQUIREMENTS NOTICE

    Binding requirements are indexed with the pharse 'shall[dd]' where dd is
    an unique number for each header section. This conforms to standard
    federal government practices, 490A (the 3.2.3.6 entry in the STD490A
    manpage). In accordance with the License, Software Diamonds is not
    liable for any requirement, binding or otherwise.

  LICENSE

    Software Diamonds permits the redistribution and use in source and
    binary forms, with or without modification, provided that the following
    conditions are met:

    1   Redistributions of source code must retain the above copyright
        notice, this list of conditions and the following disclaimer.

    2   Redistributions in binary form must reproduce the above copyright
        notice, this list of conditions and the following disclaimer in the
        documentation and/or other materials provided with the distribution.

    SOFTWARE DIAMONDS, http::www.softwarediamonds.com, PROVIDES THIS
    SOFTWARE 'AS IS' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT
    NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
    FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL SOFTWARE
    DIAMONDS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
    SPECIAL,EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
    TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,DATA, OR
    PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
    LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING USE
    OF THIS SOFTWARE, EVEN IF ADVISED OF NEGLIGENCE OR OTHERWISE) ARISING IN
    ANY WAY OUT OF THE POSSIBILITY OF SUCH DAMAGE.

SEE_ALSO:
    Math::BigInt
    Math::BigFloat
    Data::Secs2
Title Page
     Software Version Description

     for

     Data::SecsPack - pack and unpack numbers in accordance with SEMI E5-94

     Revision: B

     Version: 0.03

     Date: 2004/04/24

     Prepared for: General Public 

     Prepared by:  SoftwareDiamonds.com E<lt>support@SoftwareDiamonds.comE<gt>

     Copyright: copyright  2003 Software Diamonds

     Classification: NONE

1.0 SCOPE
    This paragraph identifies and provides an overview of the released
    files.

  1.1 Identification

    This release, identified in 3.2, is a collection of Perl modules that
    extend the capabilities of the Perl language.

  1.2 System overview

    The "Data::SecsPack" module extends the Perl language (the system).

    The subroutines in the "Data::SecsPack" module packs and unpacks numbers
    in accordance with SEMI E5-94. The E5-94 establishes the standard for
    communication between the equipment used to fabricate semiconductors and
    the host computer that controls the fabrication. The equipment in a
    semiconductor factory (fab) or any other fab contains every conceivable
    known microprocessor and operating system known to man. And there are a
    lot of specialize real-time embedded processors and speciallize
    real-time embedded operating systems in addition to the those in the PC
    world.

    The communcication between host and equipment used packed nested list
    data structures that include arrays of characters, integers and floats.
    The standard has been in place and widely used in China, Germany, Korea,
    Japan, France, Italy and the most remote places on this planent for
    decades. The SEMI E5 standard was first finalized in 1985 and has not
    changed much over the decades. The basic data structure and packed data
    formats have not changed for decades. This stands in direct
    contradiction to common conceptions of many in the Perl snail-time
    community that there is no standard for transferring such things as
    binary floats between machines little less sending nested list data as
    small compact binary.

    The "Data::Str2int" module translates an scalar string to a scalar
    integer. Perl itself has a documented function, '0+$x', that converts a
    scalar to so that its internal storage is an integer (See p.351, 3rd
    Edition of Programming Perl). If it cannot perform the conversion, it
    leaves the integer 0. Surprising not all Perls, some Microsoft Perls in
    particular, may leave the internal storage as a scalar string.

    The <str2int> function is basically the same except if it cannot perform
    the conversion to an integer, it returns an "undef" instead of a 0.
    Also, if the string is a decimal or floating point, it will return an
    undef. This makes it not only useful for forcing an integer conversion
    but also for testing a scalar to see if it is in fact an integer scalar.

  1.3 Document overview.

    This document releases Data::SecsPack version 0.03 providing a
    description of the inventory, installation instructions and other
    information necessary to utilize and track this release.

3.0 VERSION DESCRIPTION
    All file specifications in this SVD use the Unix operating system file
    specification.

  3.1 Inventory of materials released.

    This document releases the file

     Data-SecsPack-0.03.tar.gz

    found at the following repository(s):

      http://www.softwarediamonds/packages/
      http://www.perl.com/CPAN/authors/id/S/SO/SOFTDIA/

    Restrictions regarding duplication and license provisions are as
    follows:

    Copyright.
        copyright  2003 Software Diamonds

    Copyright holder contact.
         603 882-0846 E<lt>support@SoftwareDiamonds.comE<gt>

    License.
        Software Diamonds permits the redistribution and use in source and
        binary forms, with or without modification, provided that the
        following conditions are met:

        1   Redistributions of source code, modified or unmodified must
            retain the above copyright notice, this list of conditions and
            the following disclaimer.

        2   Redistributions in binary form must reproduce the above
            copyright notice, this list of conditions and the following
            disclaimer in the documentation and/or other materials provided
            with the distribution.

        SOFTWARE DIAMONDS, http://www.SoftwareDiamonds.com, PROVIDES THIS
        SOFTWARE 'AS IS' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING,
        BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
        FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL
        SOFTWARE DIAMONDS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
        SPECIAL,EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
        LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
        USE,DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
        ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
        OR TORT (INCLUDING USE OF THIS SOFTWARE, EVEN IF ADVISED OF
        NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE POSSIBILITY
        OF SUCH DAMAGE.

  3.2 Inventory of software contents

    The content of the released, compressed, archieve file, consists of the
    following files:

     file                                                         version date       comment
     ------------------------------------------------------------ ------- ---------- ------------------------
     lib/Docs/Site_SVD/Data_SecsPack.pm                           0.03    2004/04/24 revised 0.01
     MANIFEST                                                     0.03    2004/04/24 generated, replaces 0.01
     Makefile.PL                                                  0.03    2004/04/24 generated, replaces 0.01
     README                                                       0.03    2004/04/24 generated, replaces 0.01
     lib/Data/SecsPack.pm                                         0.03    2004/04/24 revised 0.02
     t/Data/SecsPack.d                                            0.02    2004/04/24 revised 0.01
     t/Data/SecsPack.pm                                           0.02    2004/04/24 revised 0.01
     t/Data/SecsPack.t                                            0.02    2004/04/24 revised 0.01
     t/Data/SecsPackStress.d                                      0.01    2004/04/24 new
     t/Data/SecsPackStress.pm                                     0.01    2004/04/24 new
     t/Data/SecsPackStress.t                                      0.01    2004/04/24 new
     t/Data/File/Package.pm                                       1.16    2004/04/24 unchanged
     t/Data/Test/Tech.pm                                          1.2     2004/04/24 revised 1.17
     t/Data/Data/Secs2.pm                                         1.17    2004/04/24 revised 1.15

  3.3 Changes

    Changes are as follows:

    Data::SecsPack 0.01
        Originated

    Data::SecsPack 0.02
        Adding support for packing and unpacking floats and flushing out to
        provide full support for packing and unpacking all SEMI E5-94 SECSII
        numeric formats.

    Data::SecsPack 0.03
        There were the same test software failures and Unit Under Test (UUT)
        failures for the following platforms:

         Subject: FAIL Data-SecsPack-0.02 ppc-darwin-thread-multi 7.2.0 
         From: nothingmuch@woobling.org 
         perl5: (revision 5.0 version 8 subversion 3) configuration:
         Platform: osname=darwin, osvers=7.2.0, archname=ppc-darwin-thread-multi

         From: "Thurn, Martin" <martin.thurn@ngc.com> 
         Subject: FAIL Data-SecsPack-0.02 sun4-solaris-thread-multi 2.8
         perl5: (revision 5.0 version 9 subversion 0)
         Platform: osname=solaris, osvers=2.8, archname=sun4-solaris-thread-multi

        The test software failures were caused by invalid data from the UUT
        failure. The UUT failures caused non-numeric data for
        "pass_fail_tolerance" in "SecsPackStress.t". Corrected
        "pass_fail_tolerance" by detecting non-numeric data and returning a
        failure. Failure of UUT "unpack_float" caused a "undef" passed to
        "binary2hex". Corrected "binary2hex" by detecting and changing
        "undef" to to "''".

        There were two groups of UUT failures.

        In the first group the test results were as the test results was two
        member array of "Math::BigInt" number hash while the test was
        expecting just two simple Perl numbers. The numbers in the
        "Data::BigInt" hash were correct just in a different format. For
        example,

         t/Data/SecsPackStress....# Test 2 got: 'U1[1] 80
         L[4]
           A[0]
           A[5] ARRAY
           L[6]
             A[12] Math::BigInt
             A[4] HASH
             A[4] sign
             A[1] +
             A[5] value
             U1[1] 128
           L[6]
             A[12] Math::BigInt
             A[4] HASH
             A[4] sign
             A[1] +
             A[5] value
             U1[1] 0
         ' (t/Data/SecsPackStress.t at line 184)
         #   Expected: 'U1[1] 80
         U1[2] 128 0
         '

        The second group of failures the UUT "unpack_float" subroutine is
        always returning a NaN. This subroutine heavily uses the native
        floating point for calcuations and it appears to be limiting out
        with "Not a Number" type error.

         # Test 27 got: '0' (t/Data/SecsPackStress.t at line 396 fail #2)
         #    Expected: '1' (got: NaN, expected: -10.5
         # actual tolerance: NaN, expected tolerance: 0.0001)

        Check version of "Math::BigInt" in the ActiveState Perl 5.06
        distribution to those on CPAN. ActiveState "Math::BigInt" was 0.01
        while CPAN was 1.70. The difference between the two is night and
        day. Install "Math::BigInt" 1.70. Tests fails the same as above.

        Corrected all "Data::SecsPack" subroutines to use new features and
        intefaces of "Math::BigInt" 1.70 including a load "die" if the
        "Math::BitInt" version is not 1.70 or greater

        Replaced as many native float operations as possible with many new
        "Math::BigInt" subroutines in Version 1.70 as possible. Removed work
        arounds for subroutines such as "rsft" do not not work completely in
        1.0 but do in 1.70. The "unpack_float" subroutine does not depend
        heavily on native float which comes in a different flavor for each
        and every site. There may be exceptions to this statement but if
        they are, they a few and far between.

  3.4 Adaptation data.

    This installation requires that the installation site has the Perl
    programming language installed. There are no other additional
    requirements or tailoring needed of configurations files, adaptation
    data or other software needed for this installation particular to any
    installation site.

  3.5 Related documents.

    There are no related documents needed for the installation and test of
    this release.

  3.6 Installation instructions.

    Instructions for installation, installation tests and installation
    support are as follows:

    Installation Instructions.
        To installed the release file, use the CPAN module pr PPM module in
        the Perl release or the INSTALL.PL script at the following web site:

         http://packages.SoftwareDiamonds.com

        Follow the instructions for the the chosen installation software.

        If all else fails, the file may be manually installed. Enter one of
        the following repositories in a web browser:

          http://www.softwarediamonds/packages/
          http://www.perl.com/CPAN/authors/id/S/SO/SOFTDIA/

        Right click on 'Data-SecsPack-0.03.tar.gz' and download to a
        temporary installation directory. Enter the following where $make is
        'nmake' for microsoft windows; otherwise 'make'.

         gunzip Data-SecsPack-0.03.tar.gz
         tar -xf Data-SecsPack-0.03.tar
         perl Makefile.PL
         $make test
         $make install

        On Microsoft operating system, nmake, tar, and gunzip must be in the
        exeuction path. If tar and gunzip are not install, download and
        install unxutils from

         http://packages.softwarediamonds.com

    Prerequistes.
           "Math::BigInt" => "1.50"

    Security, privacy, or safety precautions.
        None.

    Installation Tests.
        Most Perl installation software will run the following test
        script(s) as part of the installation:

         t/Data/SecsPack.t
         t/Data/SecsPackStress.t

    Installation support.
        If there are installation problems or questions with the
        installation contact

         603 882-0846 E<lt>support@SoftwareDiamonds.comE<gt>

  3.7 Possible problems and known errors

    There is still much work needed to ensure the quality of this module as
    follows:

    *   Increase the accuracy of packing floats with large exponents and
        unpacking floats.

    *   State the functional requirements for each method including not only
        the GO paths but also what to expect for the NOGO paths

    *   All the tests are GO path tests. Should add NOGO tests.

    *   Add the requirements addressed as *# R: * comment to the tests

4.0 NOTES
    The following are useful acronyms:

    .d  extension for a Perl demo script file

    .pm extension for a Perl Library Module

    .t  extension for a Perl test script file

2.0 SEE ALSO
    Data::SecsPack
    Docs::US_DOD::SVD
