##### # 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, C<@options>, that subroutine will also process an array reference, C<\@options>, C<[@options]>, or hash reference, C<\%options>, C<{@options}>. If a subroutine will process an array reference, C<\@options>, C<[@options]>, that subroutine will also process a hash reference, C<\%options>, C<{@options}>. See the description for a subroutine for details and exceptions.
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 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:
``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 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 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 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:
=back
For example, to find the IEEE 754-1985 float of -10.5
01010000000000000000000
130 decimal converted to 8 bit binary
10000010Combining 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.
$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.
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:
$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.
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
$Data::SecsPack::default_options->{version};
The options are as follows:
used by values default subroutine option value 1st ---------------------------------------------------------- big_float_version \d+\.\d+ big_int_version \d+\.\d+ version \d+\.\d+
warnings 0 1 die 0 1
bytes2int
float2binary decimal_integer_digits 20 \d+ extra_decimal_fraction_digits 5 \d+ decimal_fraction_digits binary_fraction_bytes
ifloat2binary decimal_fraction_digits 25 \d+ binary_fraction_bytes 10 \d+
int2bytes pack_float decimal_integer_digits extra_decimal_fraction_digits decimal_fraction_digits binary_fraction_bytes
pack_int
pack_num nomix 0 1 decimal_integer_digits extra_decimal_fraction_digits decimal_fraction_digits binary_fraction_bytes
str2float ascii_float 0 1 str2int unpack_float unpack_int unpack_num
For options with a default value and subroutine, see the subroutine for a description of the option. Each subroutine that uses an option or uses a subroutine that uses an option has an option input. The option input overrides the startup option from the <Data::SecsPack> object.
The description of the options without a subroutine are as follows:
option description -------------------------------------------------------------- big_float_version Math::BigFloat version big_int_version Math::BigInt version version Data::SecsPack version
warnings issue a warning on subroutine events die die on subroutine events
They really versions should not be changed unless the intend is to provided fraudulent versions.
($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
.
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.
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.
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 $iexponent
breakout to adjust the ifloat2binary
subroutine results using
native float arith.
If the float2binary
subroutine encounters an event where it cannot
continue, it halts processing, and returns the
event as
(undef,$event)
The events are as follows:
"No inputs\n\tData::SecsPack::float2binary-1\n" The C<float2binary> also passes on any C<ifloat2binary> events. Check the C<$binary_magnitude> for an C<undef>, to see if the subroutine cannot process the decimal exponent.
($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 $binary_magnitude
is 1 and the rest of the bytes are
a base 2 fraction.
Since all the calculations use basic integer arith, there are
practical limits on the computer resources. Basically the limit is that
with a zero exponent, the decimal point is within the significant
imagnitude
digits. Within these limitations, the accuracy, by
chosen large enough limits for the binary fraction, is perfect.
If the ifloat2binary
subroutine encounters an event where it cannot
continue, it halts processing, and returns the
event as
(undef,$event)
The events are as follows:
"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
.
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
@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.
$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
$secspack
method will perform that subroutine
with the options saved in secspack
.
($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 $integers
.
If the pack_float
subroutine encounters an event where it cannot
continue, it halts processing, and returns the
event as
(undef,$event)
The events are as follows:
"No inputs.\n\tData::SecsPack::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 $integers
.
If the pack_int
subroutine encounters an event where it cannot
continue, it halts processing, and returns the
event as
(undef,$event)
The events are as follows:
"No inputs.\n\tData::SecsPack::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
.
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
$format
,
if the nomix
option is set, the pack_num
subroutine will
pack all leading, integers, decimals and floats as multicell float
with the smallest space; otherwise, it will stop at the first
decimal or float encountered and just pack the integers.
The pack_num
subroutine processes @strings
in two steps.
In the first step, the
pack_num
subroutine uses str2int
and/or str2float
subroutines to parse the leading
numbers from the @strings
as follows:
([@strings], @integers) = str2int(@strings); ([@strings], @floats) = str2float(@strings);
In the second step,
the pack_num
subroutine uses pack_int
and/or pacK_float
to pack the parsed numbers.
If the pack_nym
subroutine encounters an event where it cannot
continue, it halts processing, and returns the
event as
(undef,$event)
The events are as follows:
"No inputs.\n\tData::SecsPack::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});
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 floats @floats
.
For the ascii_float
option, the members of the @floats
are scalar
strings of the float numbers; otherwise, the members are a reference
to an array of [$decimal_magnitude, $decimal_exponent]
where the decimal
point is set so that there is one decimal digit to the left of the decimal
point for $decimal_magnitude.
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.
$integer = str2int($string); $integer = str2int($string, [@options]); $integer = str2int($string, {@options});
(\@strings, @integers) = str2int(@strings); (\@strings, @integers) = str2int(@strings, [@options]); (\@strings, @integers) = str2int(@strings, {@options});
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.
What is $x
for the following:
my $x = 0 + '0x100'; # $x is 0 with a warning
Instead use str2int
uses a few simple Perl lines, without
any evals
starting up whatevers or firing up the
regular expression engine with its interpretative overhead,
to provide a slightly different response as follows:>.
$x = str2int('033'); # $x is 27 $x = str2int('0xFF'); # $x is 255 $x = str2int('255'); # $x is 255 $x = str2int('hello'); # $x is undef no warning $x = str2int(0.5); # $x is undef no warning $x = str2int(1E0); # $x is 1 $x = str2int(0xf); # $x is 15 $x = str2int(1E30); # $x is undef no warning
The scalar str2int
subroutine performs the conversion to an integer
for strings that look like integers and actual integers without
generating warnings.
A non-numeric string, decimal or floating string returns an ``undef''
instead of the 0 and a warning
that 0+'hello'
produces.
This makes it not only useful for forcing an integer conversion but
also for testing a scalar to see if it is in fact an integer scalar.
The scalar str2int
is the same and supercedes C&<Data::StrInt::str2int>.
The Data::SecsPack
program module superceds the Data::StrInt
program module.
The str2int
subroutine, in an array context, supports converting multiple run of
integers in an array of strings @strings
to an array of integers, @integers
.
It keeps converting the strings, starting with the first string in @strings
,
continuing to the next and next until it fails a conversion.
The str2int
returns the remaining string data in @strings
and
the array of integers @integers
.
\@floats = unpack_float($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
.
If the unpack_float
subroutine encounters an event where it cannot
continue, it halts processing, and returns the
event as
$event
The events are as follows:
"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 $event
. To determine a valid return or an error,
check that ref
of the return exists or is 'ARRAY
'.
=head2 unpack_int
\@integers = unpack_int($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
.
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
'.
If the unpack_float
subroutine encounters an event where it cannot
continue, it halts processing, and returns the
event as
$event
The events are as follows:
"No inputs\ntData::SecsPack::unpack_int-1\n" "Format $format_in not supported.\n"tData::SecsPack::unpack_int-2\n"
The unpack_num
subroutine, thus, returns a reference, \@floats
, to the unpacked float array
or scalar error message $event
. To determine a valid return or an error,
check that ref
of the return exists or is 'ARRAY
'.
\@numbers = unpack_num($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
.
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
'.
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 #
$unpack_numbers
# [ # '7.800000000000017486E1', # '4.500000000000006245E0', # '2.5E-1', # '6.4500000000000376452E10' # ] #
Running the test script C<SecsPack.t> and C<SecsPackStress.t> verifies the requirements for this module. The C<tmake.pl> cover script for C<Test::STDmaker|Test::STDmaker> automatically generated the C<SecsPack.t> and C<SecsPackStress.t> test scripts, the C<SecsPack.d> and C<SecsPackStress.d> demo scripts, and the C<t::Data::SecsPack> and C<t::Data::SecsPackStress> STD program module PODs, from the C<t::Data::SecsPack> and C<t::Data::SecsPackStress> program module's content. The C<t::Data::SecsPack> and C<t::Data::SecsPackStress> program modules are in the distribution file F<Data-SecsPack-$VERSION.tar.gz>.
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:
=back
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.