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 "$binary_magnitude" is 1 and the rest of the bytes are
a base 2 fraction.
Since all the calculations use basic integer arith, there are practical
limits on the computer resources. Basically the limit is that with a
zero exponent, the decimal point is within the significant "imagnitude"
digits. Within these limitations, the accuracy, by chosen large enough
limits for the binary fraction, is perfect.
The first step of the "ifloat2binary" subroutine is zero out "iexponent"
by breaking up the "imagnitude" into an integer part "integer" and
fractional part "fraction" consist with the "iexponent". The
c will add as many significant decimal zeros to the right
of "integer" in order to zero out "iexponent"; likewise it will add as
many decimal zeros to the left of "integer" to zero out "exponent"
within the limit set by the option "decimal_fraction_digits". If
"ifloat2binary" cannot zero out "iexponent" without violating the
"decimal_fraction_digits", "ifloat2binary" will discontinue processing
and return an "undef" "$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+$x', that converts a scalar to so that its internal storage
is an integer (See p.351, 3rd Edition of Programming Perl). If it cannot
perform the conversion, it leaves the integer 0. Surprising not all
Perls, some Microsoft Perls in particular, may leave the internal
storage as a scalar string.
The scalar "str2int" subroutine is basically the same except if it
cannot perform the conversion to an integer, it returns an "undef"
instead of a 0. Also, if the string is a decimal or floating point, it
will return an undef. This makes it not only useful for forcing an
integer conversion but also for testing a scalar to see if it is in fact
an integer scalar. The scalar "str2int" is the same and supercedes
C&. The "Data::SecsPack" program module
superceds the "Data::SecsPack" program module.
The "str2int" subroutine, in an array context, supports converting
multiple run of integers in an array of strings "@strings" to an array
of integers, "@integers". It keeps converting the strings, starting with
the first string in "@strings", continuing to the next and next until it
fails a conversion. The "str2int" returns the remaining string data in
"@strings" and the array of integers "@integers".
unpack_float
\@floats = unpack_float($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
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 Esupport@SoftwareDiamonds.comE
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 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 Esupport@SoftwareDiamonds.comE
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"
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 Esupport@SoftwareDiamonds.comE
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