Data::SecsPack - pack and unpack numbers in accordance with SEMI E5-94
##### # 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, $binary_exponent) = ifloat2binary($imagnitude, $iexponent, @options); @bytes = int2bytes( $big_integer ); ($format, $floats) = pack_float($format, @string_floats, [@options]); ($format, $integers) = pack_int($format, @string_integers, [@options]); ($format, $numbers, @string) = pack_num($format, @strings, [@options]); $float = str2float($string, [@options]); (\@strings, @floats) = str2float(@strings, [@options]); $integer = str2int($string, [@options]); (\@strings, @integers) = str2int(@strings, [@options]); \@ingegers = unpack_int($format, $integer_string, @options); \@floats = unpack_float($format, $float_string, @options); \@numbers = unpack_num($format, $number_string), @options; ##### # Class, Object interface # # For class interface, use Data::SecsPack instead of $self # use Data::SecsPack; $secspack = 'Data::SecsPack'; # uses built-in config object $secspack = new Data::SecsPack(@options); $big_integer = bytes2int( @bytes ); ($binary_magnitude, $binary_exponent) = $secspack->float2binary($magnitude, $exponent, @options); ($binary_magnitude, $binary_exponent) = $secspack->ifloat2binary($imagnitude, $iexponent, @options); @bytes = $secspack->int2bytes( $big_integer ); ($format, $floats) = $secspack->pack_float($format, @string_integers, [@options]); ($format, $integers) = $secspack->pack_int($format, @string_integers, [@options]); ($format, $numbers, @strings) = $secspack->pack_num($format, @strings, [@options]); $integer = $secspack->str2int($string, [@options]) (\@strings, @integers) = $secspack->str2int(@strings, [@options]); $float = $secspack->str2float($string, [@options]); (\@strings, @floats) = $secspack->str2float(@strings, [@options]); \@ingegers = $secspack->unpack_int($format, $integer_string, @options); \@floats = $secspack->unpack_float($format, $float_string, @options); \@numbers = $secspack->unpack_num($format, $number_string, @options);
Generally, if a subroutine will process a list of options, @options, that subroutine will also process an array reference, \@options, [@options], or hash reference, \%options, {@options}. If a subroutine will process an array reference, \@options, [@options], that subroutine will also process a hash reference, \%options, {@options}. See the description for a subroutine for details and exceptions.
@options
\@options
[@options]
\%options
{@options}
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.
Data::SecsPack
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 corners on this planent for decades. The basic data structure and packed data formats have not changed for decades.
This stands in direct contradiction to the common conceptions of many in the Perl community and most other communities. The following quote is taken from page 761, Programming Perl third edition, discussing the pack subroutine:
pack
"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."
There are a lot of things that go over the net that have industry or military standards but no RFCs. So unless you dig them out, you will never know they exist. While RFC and military standards may be freely copyied, industry standards are usually copyrighted. This means if you want to read the standard, you have to pay whatever the market bears. ISO standards, SEMI stardards, American National Standards, IEEE standards beside being boring are expensive. In other words, you do not see them flying out the door at the local Barnes and Nobles. In fact, you will not even find them inside the door.
It very easy to run these non RFC standard protocols over the net. Out of 64,000 ports, pick a port of opportunity (hopefully not one of those low RFC preassigned ports) and configure the equipment and host to the same IP and port. Many times the software will allow a remote console that is watch only. The watch console may even be a web server on port 80. If there is a remote soft console, you can call up or e-mail the equipment manufacturer's engineer in say Glouster, MA, USA and tell him the IP and port so he can watch his manchine mangle a cassette of wafers with a potential retail value of half million dollars.
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 endian-ness of SEMI E5-94 is the first MSB byte, floats sign bit first. Maybe this is because it makes it easy to spot numbers in a packed data structure.
The nested data has many performance advantages over the common SQL culture of viewing and representing data as tables. The automated fabs of the world make use of SEMI E5-94 nested data not only for real-time communication (TCP/IP RS-2332 etc) between machines but also for snail-time processing as such things as logs and performance data.
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 to test software 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 with your favorite or not so favorite co-worker just to get into the clean room. And make sure not to do anything that will scatch a wafer with a lot of Intel Pentiums on them. It is totally amazing that the product does get out the door.
The Data::SecsPack suroutines packs and unpacks numbers in accordance with SEMI, http://www.semi.org, 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.semi.org
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,
the Most Significat Byte (MSB) of numbers for formats S2, S4, S8, U2, U4, U8 is sent first
the signed bit for formats F4 and F8 are sent first.
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:
The sign is one bit, 0 for positive and 1 for negative.
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.
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 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.
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
$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.
bytes2int
@bytes
$big_integer
Data::BigInt
$old_value = config( $option ); $old_value = config( $option => $new_value); (@all_options) = config( );
When Perl loads the Data::SecsPack program module, Perl creates the Data::SecsPack subroutine Data::SecsPack object $Data::SecsPack::subroutine_secs using the new method. Using the config subroutine writes and reads the $Data::SecsPack::subroutine_secs object.
$Data::SecsPack::subroutine_secs
new
config
Using the config as a class method,
Data::SecsPack->config( @_ )
also writes and reads the $Data::SecsPack::subroutine_secs object.
Using the config as an object method writes and reads that object.
The Data:SecsPack subroutines used as methods for that object will use the object underlying data for their startup (default options) instead of the $Data::SecsPack::subroutine_secs object. It goes without saying that that object should have been created using one of the following:
Data:SecsPack
$object = $class->Data::SecsPack::new(@_) $object = Data::SecsPack::new(@_) $object = new Data::SecsPack(@_)
The underlying object data for the Data::SecsPack options defaults is the class Data::Startup object $Data::SecsPack::default_options. For object oriented conservative purist, the config subroutine is the accessor function for the underlying object hash.
Data::Startup
$Data::SecsPack::default_options
Since the data are all options whose names and usage is frozen as part of the Data::SecsPack interface, the more liberal minded, may avoid the config accessor function layer, and access the object data directly by a statement such as
$Data::SecsPack::default_options->{version};
The options are as follows:
used by values default subroutine option value 1st ---------------------------------------------------------- big_float_version \d+\.\d+ big_int_version \d+\.\d+ version \d+\.\d+ warnings 0 1 die 0 1 bytes2int float2binary decimal_integer_digits 20 \d+ extra_decimal_fraction_digits 5 \d+ decimal_fraction_digits binary_fraction_bytes ifloat2binary decimal_fraction_digits 25 \d+ binary_fraction_bytes 10 \d+ int2bytes pack_float decimal_integer_digits extra_decimal_fraction_digits decimal_fraction_digits binary_fraction_bytes pack_int pack_num nomix 0 1 decimal_integer_digits extra_decimal_fraction_digits decimal_fraction_digits binary_fraction_bytes unpack_float unpack_int unpack_num
For options with a default value and subroutine, see the subroutine for a description of the option. Each subroutine that uses an option or uses a subroutine that uses an option has an option input. The option input overrides the startup option from the <Data::SecsPack> object.
The description of the options without a subroutine are as follows:
option description -------------------------------------------------------------- big_float_version Math::BigFloat version big_int_version Math::BigInt version version Data::SecsPack version warnings issue a warning on subroutine events die die on subroutine events
They really versions should not be changed unless the intend is to provided fraudulent versions.
($binary_magnitude, $binary_exponent) = float2binary($magnitude, $exponent); ($binary_magnitude, $binary_exponent) = float2binary($magnitude, $exponent, @options); ($binary_magnitude, $binary_exponent) = float2binary($magnitude, $exponent, [@options]); ($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.
ifloat2binary
$magnitude
$exponent
$binary_magnitude
$binary_exponent
The ifloat2binary assumes that the decimal point is set by iexponent 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.
iexponent
imagnitude
The float2binary subroutine uses the ifloat2binary for the small $exponent part and the Math::BigFloat subroutines to correct the ifloat2binary for the remaing exponent factor outside the range of the ifloat2binary subroutine.
float2binary
Math::BigFloat
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.
decimal_integer_digits
$decial_fraction_digits
extra_decimal_fraction_digits
$iexponent
$ixponent
If the float2binary subroutine encounters an event where it cannot continue, it halts processing, and returns the event as
(undef,$event)
The events are as follows:
"No inputs\n\tData::SecsPack::float2binary-1\n"
The float2binary also passes on any ifloat2binary events. Check the $binary_magnitude for an undef, to see if the subroutine cannot process the decimal exponent.
undef
($binary_magnitude, $binary_exponent) = ifloat2binary($imagnitude, $iexponent); ($binary_magnitude, $binary_exponent) = ifloat2binary($imagnitude, $iexponent, @options); ($binary_magnitude, $binary_exponent) = ifloat2binary($imagnitude, $iexponent, [@options]); ($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 iexponent 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.
$ifloat2binary
$imagnitude
Math::BigInt
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.
If the ifloat2binary subroutine encounters an event where it cannot continue, it halts processing, and returns the event as
"No inputs\n\tData::SecsPack::ifloat2binary-1\n" "The exponent, $exponent, is out of range for $magnitude.\n\tData::SecsPack::ifloat2binary-2\n"
Check the $binary_magnitude for an undef, to see if the subroutine cannot process the decimal exponent.
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.
integer
fraction
exponent
decimal_fraction_digits
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.
binary_fraction_bytes
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
@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.
int2bytes
Data:BigInt
$bit_integer
$bytes[0]
$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.
$secspack-
$secspack
secspack
($format, $floats) = pack_float($format, @string_integers); ($format, $floats) = pack_float($format, @string_integers, [@options]); ($format, $floats) = pack_float($format, @string_integersm {@options});
The pack_float 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.
pack_float
Item Format Code Table
$format
SEMI E5-94
pack_int
F
$integers
If the pack_float subroutine encounters an event where it cannot continue, it halts processing, and returns the event as
"No inputs.\n\tData::SecsPack::pack_float-1\n" "Format $format is not a floating point format.\n\tData::SecsPack::pack_float-2\n" "F4 exponent overflow.\n\tData::SecsPack::pack_float-3\n" "F4 xponent underflow.\n\tData::SecsPack::pack_float-4\n" "F8 exponent overflow.\n\tData::SecsPack::pack_float-5\n" "F8 xponent underflow.\n\tData::SecsPack::pack_float-6\n"
The float2binary also passes on any float2binary and ifloat2binary events. Check the $format for an undef, to see if the subroutine cannot continue processing.
($format, $integers) = pack_int($format, @string_integers); ($format, $integers) = pack_int($format, @string_integers, [@options]); ($format, $integers) = pack_int($format, @string_integers, {options});
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.
I I1 I2 I8
If the pack_int subroutine encounters an event where it cannot continue, it halts processing, and returns the event as
"No inputs.\n\tData::SecsPack::pack_int-1\n" "Format $format is not an integer format.\ntData::SecsPack::pack_int-2\n" "No integers in the input.\ntData::SecsPack::pack_int-3\n" "Signed number encountered when unsigned specified.\ntData::SecsPack::pack_int-4\n" "Integer bigger than format length of $max_bytes bytes.\ntData::SecsPack::pack_int-5\n"
Check the $format for an undef, to see if the subroutine cannot continue processing.
($format, $numbers, @strings) = pack_num($format, @strings); ($format, $numbers, @strings) = pack_num($format, @strings, [@options]); ($format, $numbers, @strings) = pack_num($format, @strings, {@options});
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.
pack_num
@strings
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.
I I1 I2 I4 F
Item FOrmat Code Table
For the I $format, if the nomix option is set, 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.
I
nomix
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:
str2int
str2float
([@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.
pacK_float
If the pack_nym subroutine encounters an event where it cannot continue, it halts processing, and returns the event as
pack_nym
"No inputs.\n\tData::SecsPack::pack_num-1\n" "Format $format is not an integer or floating point format.\ntData::SecsPack::pack_num-2\n" "No numbers in the input.\ntData::SecsPack::pack_num-3\n"
The float2binary also passes on any float2binary ifloat2binary pack_int pack_float events. Check the $format for an undef, to see if the subroutine cannot continue processing.
$float = str2float($string); $float = str2float($string, [@options]); $float = str2float($string, {@options}); (\@strings, @floats) = str2float(@strings); (\@strings, @floats) = str2float(@strings, [@options]); (\@strings, @floats) = str2float(@strings, {@options});
Obsoleted and superceded by the Data::Str2Num-str2float> subroutine in the Data::Str2Num package.
Data::Str2Num-
$integer = str2int($string); $integer = str2int($string, [@options]); $integer = str2int($string, {@options}); (\@strings, @integers) = str2int(@strings); (\@strings, @integers) = str2int(@strings, [@options]); (\@strings, @integers) = str2int(@strings, {@options});
Obsoleted and superceded by the Data::Str2Num-str2integer> subroutine in the Data::Str2Num package.
\@floats = unpack_float($format, $float_string); \@floats = unpack_float($format, $float_string, @options); \@floats = unpack_float($format, $float_string, [@options]); \@floats = unpack_float($format, $float_string, {@options});
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.
unpack_num
$float_string
F4 F8
If the unpack_float subroutine encounters an event where it cannot continue, it halts processing, and returns the event as
unpack_float
$event
"No inputs\ntData::SecsPack::unpack_float-1\n" "Format $format_in not supported.\n"tData::SecsPack::unpack_float-2\n"
The unpack_num subroutine, thus, returns a reference, \@floats, to the unpacked float array or scalar error message $event. To determine a valid return or an error, check that ref of the return exists or is 'ARRAY'.
\@floats
ref
ARRAY
\@integers = unpack_int($format, $integer_string); \@integers = unpack_int($format, $integer_string, @options); \@integers = unpack_int($format, $integer_string, [@options]); \@integers = unpack_int($format, $integer_string, {@options});
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.
$string_numbers
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'.
\@integers
$error
"No inputs\ntData::SecsPack::unpack_int-1\n" "Format $format_in not supported.\n"tData::SecsPack::unpack_int-2\n"
\@numbers = unpack_num($format, $number_string); \@numbers = unpack_num($format, $number_string, @options); \@numbers = unpack_num($format, $number_string, [@options]); \@numbers = unpack_num($format, $number_string, {@options});
The unpack_num subroutine unpacks an array of numbers $number_string packed in accordance with SEMI E5-94 $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.
$number_string
S1 S2 S4 U1 U2 U4 F4 F8 T
unpack_int
The pack_num subroutine does not generate any events but the subroutine does pass on any pack_int and pack_float events, returning them as a string. The unpack_num subroutine, thus, returns a reference, \@numbers, to the unpacked number array or scalar error message $event. To determine a valid return or an error, check that ref of the return exists or is 'ARRAY'.
\@numbers
Coming.
######### # perl SecsPack.d ###
~~~~~~ Demonstration overview ~~~~~
The results from executing the Perl Code follow on the next lines as comments. For example,
2 + 2 # 4
~~~~~~ The demonstration follows ~~~~~
use File::Package; my $fp = 'File::Package'; my $uut = 'Data::SecsPack'; my $loaded; ##### # Provide a scalar or array context. # my ($result,@result); ################## # 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('0xFF') # $result = $uut->str2int('0xFF') # '255' # ################## # str2int('255') # $result = $uut->str2int('255') # '255' # ################## # str2int('hello') # $result = $uut->str2int('hello') # undef # ################## # str2int(1E20) # $result = $uut->str2int(1E20) # undef # ################## # str2int(' 78 45 25', ' 512E4 1024 hello world') @numbers # my ($strings, @numbers) = str2int(' 78 45 25', ' 512E4 1024 hello world') [@numbers] # [ # '78', # '45', # '25' # ] # ################## # str2int(' 78 45 25', ' 512E4 1024 hello world') @strings # join( ' ', @$strings) # '512E4 1024 hello world' # ################## # str2float(' 78 -2.4E-6 0.0025 0', ' 512E4 hello world') numbers # ($strings, @numbers) = str2float(' 78 -2.4E-6 0.0025 0', ' 512E4 hello world') [@numbers] # [ # [ # '78', # '1' # ], # [ # '-24', # '-6' # ], # [ # '25', # -3 # ], # [ # '0', # -1 # ], # [ # '512', # '6' # ] # ] # ################## # str2float(' 78 -2.4E-6 0.0025 0', ' 512E4 hello world') @strings # join( ' ', @$strings) # 'hello world' # ################## # str2float(' 78 -2.4E-6 0.0025 0xFF 077 0', ' 512E4 hello world', {ascii_float => 1}) numbers # ($strings, @numbers) = str2float(' 78 -2.4E-6 0.0025 0xFF 077 0', ' 512E4 hello world', {ascii_float => 1}) [@numbers] # [ # '78', # '-2.4E-6', # '0.0025', # '255', # '63', # '0', # '512E4' # ] # ################## # str2float(' 78 -2.4E-6 0.0025 0xFF 077 0', ' 512E4 hello world', {ascii_float => 1}) @strings # 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(I, 78 45 25 512 1024 100000 hello world) format # $format # 'U4' # ################## # pack_num(I, 78 45 25 512 1024 100000 hello world) numbers # unpack('H*',$numbers) # '0000004e0000002d000000190000020000000400000186a0' # ################## # pack_num(I, 78 45 25 512 1024 100000 hello world) @strings # [@strings] # [ # 'hello world' # ] # ################## # unpack_num(U4, 78 45 25 512 1024 100000 hello world) error check # ref(my $unpack_numbers = unpack_num($expected_format,$numbers)) # 'ARRAY' # ################## # unpack_num(U4, 78 45 25 512 1024 100000 hello world) 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'; $expected_strings = ['hello world']; my @expected_unpack = ( '7.800000000000017486E1', '4.500000000000006245E0', '2.5E-1', '6.4500000000000376452E10' ); ($format, $numbers, @strings) = pack_num('I',@test_strings); ################## # pack_num(I, 78 4.5 .25 6.45E10 hello world) format # $format # 'F8' # ################## # pack_num(I, 78 4.5 .25 6.45E10 hello world) numbers # unpack('H*',$numbers) # '405380000000000040120000000000003fd0000000000000422e08ffca000000' # ################## # pack_num(I, 78 4.5 .25 6.45E10 hello world) @strings # [@strings] # [ # 'hello world' # ] # ################## # unpack_num(F8, 78 4.5 .25 6.45E10 hello world) error check # ref($unpack_numbers = unpack_num($expected_format,$numbers)) # 'ARRAY' # ################## # unpack_num(F8, 78 4.5 .25 6.45E10 hello world) numbers # $unpack_numbers # [ # '7.800000000000017486E1', # '4.500000000000006245E0', # '2.5E-1', # '6.4500000000000376452E10' # ] #
Running the test script SecsPack.t and SecsPackStress.t verifies the requirements for this module. The tmake.pl cover script for Test::STDmaker|Test::STDmaker automatically generated the SecsPack.t and SecsPackStress.t test scripts, the SecsPack.d and SecsPackStress.d demo scripts, and the t::Data::SecsPack and t::Data::SecsPackStress STD program module PODs, from the t::Data::SecsPack and t::Data::SecsPackStress program module's content. The t::Data::SecsPack and t::Data::SecsPackStress program modules are in the distribution file Data-SecsPack-$VERSION.tar.gz.
SecsPack.t
SecsPackStress.t
tmake.pl
Test::STDmaker|Test::STDmaker
SecsPack.d
SecsPackStress.d
t::Data::SecsPack
t::Data::SecsPackStress
The holder of the copyright and maintainer is
<support@SoftwareDiamonds.com>
Copyrighted (c) 2002 Software Diamonds
All Rights Reserved
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, STD490A 3.2.3.6. In accordance with the License, Software Diamonds is not liable for any requirement, binding or otherwise.
Software Diamonds permits the redistribution and use in source and binary forms, with or without modification, provided that the following conditions are met:
Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
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.
Commercial installation of the binary or source must visually present to the installer the above copyright notice, this list of conditions intact, that the original source is available at http://softwarediamonds.com and provide means for the installer to actively accept the list of conditions; otherwise, a license fee must be paid to Softwareware Diamonds.
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.
To install Data::SecsPack, copy and paste the appropriate command in to your terminal.
cpanm
cpanm Data::SecsPack
CPAN shell
perl -MCPAN -e shell install Data::SecsPack
For more information on module installation, please visit the detailed CPAN module installation guide.