MESSAGE
| DATE | 2010-02-02 |
| FROM | Ruben Safir
|
| SUBJECT | Subject: [NYLXS - HANGOUT] C++ Workshop I datatypes cont..
|
We discussed that everything in our computer is data. Lets expand
on this concept and create some sample programs that demonstrate this
concept and flesh out the data types in more detail as they specifically
relate to C++ programming. C++ inherits much of its core datatypes from
C. And this has been a general problem for C++ texts because it seems
that the authors of most of these books simply don't want to take the
necessary time discuss the data types as inherited from C, and just flub
it, with the assumption that the reader is familiar with the C language
specifications. I'd like to correct that, and give a more uniform and
coherent look at the C++ datatypes, which in reality, is completely
necessary for understand all the more advanced feature of C++ and object
design.
In C++ there are three categries of datatypes. There is the built in
datatypes and their built in agregates. There is standard C++ extended
Class based datatypes. And then there are user defined or created
datatypes.
A) Built in Datatypes and their aggregates
B) Standard Library C++ datatypes
C) User Created Data types.
Previously I discussed that when it comes to a computer, and any
programming language, everything is data. The code you write is data,
the information your program retrieves and manipulates as a target is
data, and the addressing of the storage locations of your data is data.
In order to think about this in a simpilfied way, we often view a
computer as a process manipulation system. This mathmatically and
theoretically involves set theory and Turing Machine priciples, which I
am not going to cover, but try to remember that when dealing with your
computer, and when programming your computer, that is system exists in
the real world. It is easy to forget this, but strangley enough,
computers aren't just good at virtual reality, but actually work in
reality...period. It is a machine. And as such it has inherent
properties, many of which we take for granted in the real world,
although
scientists and engineers don't take it for granted, but when we interact
with our digital systems, we forget, and then we suffer from poor coding
habits and computers that perplex us.
As a real world machine, computers deal with processes. And these
processes we try to control and shape to our needs. All processes, and
this isn't negotiable, but a matter of fact, have three basic elements.
They have input. They have output, and they have side affects. In the
physical world, for example, all processes result in an increase in
heat. The input is usually some kind of fuel. And the side affect is
pretty much what we find useful. A simple machine like a gasoline
powered
automobie takes in gas, air and electricity, outputs burnt fuel, gases
like Carbon Dioxide, and HEAT, and the side affect is controlled
movement.
Computers and coding is a little more complex, but the principles remain
intact. Your data, in the form of code, is sent into the CPU, along
with other inputs and data, the CPU processes that information and
returns its "state" and all the pretty pictures, sounds, printed
materials, whatever your hardware allows, is the side affect.
Now that is very internal to your system, but another useful process to
examine and to be aware of when working with your computer is process of
how information flows through your computer, and involves the end user.
On a standard computer model, the real world of process control is built
into your digital computer. Information (data) enters your computer
from the STANDARD INPUT DEVICE (your keyboard), does some work and
displays results on your STARNARD OUTPUT DEVICE. And the side affects
can be varried. But this is an artificial contruction, otherwise the
computer would be useless, because the interaction of your computer with
the real physical world would would be incomprehensible by the stupid
humans that are using it. So while for your CPU, creating recognizable
images on the screen is really a side affect, from an information theory
perspective, and from a pracital programming perspective, it is the
output. And most of your job as a programmer is to bridge that gap
between your CPU, who's output is a return of "state" information, to
get all of your programs side affects working so that the information
processing generates correct and usable information responses for the
humans communicating with it.
Now if I haven't yet confused you (and go back and reread this if I
did), data must be understood by your computer and by you. The data
that goes into your program must be understood by your computer and the
data that is outputed must be understood by you. And the side affects
of this process have to be correct as well. In fact, more often than
not, the side affects is everything because the only information your
program might be returning to you is, "OK - I'm finished - awaiting more
instructions". We'll return to these principles over and over again.
In order to help bridge the human digital divide, so to speak, your
computer handles data, and in modern machines, that data is based on the
8 bit Byte. The similist computer that individuals come in contact with
is the simple light switch. It does everything your computer does. You
have a simple switch with a bulb attached to it. Your turn the switch
on (stadard input). Your switch allows electricity to flow to the buld
and creates light and heat (side affect), and the switch records its
"state", which is ON...otherwise it wouldn't be very useful. And as
long as the system is working, it retains its state until you give it
more information (turn it off in this case). Now in this system, you
have one bit of information that can be processed. The light is either
ON or OFF. there is nothing else you can do. But if we have 2 switches
connected to a light bulb, we can engineer a system with up to 4 states.
Off, On Dim, On Brighter, On Brightest, each one representing a
different "state". With eight switches, you can represent 256
states...and THAT is exactly what yor digital computer does. Your
hardware and program determines what each of those 256 states can mean.
If you need more that 256 states to implement a side affect, lets say we
have a byte in our hardware that makes out stardard output device a
color. 256 colors is nice. The billions of colors that the human eye
can
distiguish is nicer. But then we need to stirng together multiple bytes
to implement that sceem. Perhaps, two or even 4 bytes. The same holds
true for numbers. If we are using one byte, that will only allow us to
describe 256 numbers, and in this case, to allow for positive and
negitive numbers, one of the bits in our byte usually identifies the
number as positive or negative. -127 to 127. That is called a signed
one byte integer. And unsigned integer can't represent negitive
numbers, (one bit in the byte is not used to describe negitive or
positive state) and it can represent 0-255.
Now that isn't good enough for banking, science, and most other
applications, although it is fine if we are making a card game or a dice
game, so integers are most often represents by at least 2 bytes, and
often 4 or more. And the ability to do with has to be available to the
hardware and written into the language libaries.
So what data types does C++ give us by default? Well, not colors, we
would have to create that. It gives us the following default data
types, which I'll try to explain.
char - Characters. A representation of standard characters as a single
byte, which is more than adequate for latin based languages like
English. A data input that is a character is stored in a single byte
which C++ will automatically tranlate into characters, like A-Za-z and
others as defined in ASCI.
http://www.asciitable.com/
As computers got larger, the ASCII set was extended to into 2 bytes
(ASCII extened) but in C++ a char is a SINGLE BYTE.
int - A one word Integer. Storage is hardware dependent, most normally
in something called a " word" which on the standard 32 bit hardware and
operating systems is 4 bytes (hence 32 bits). Machines are increasingly
64 bite word based hardware and the operating systems are catching up I
would asume that those systems have 8 bytes ints.
Before going forward, it is worth stating that there is a standard for
the C programming language called the ANSI C standard. You have to buy
the standard if you want it and in it has rules for how to define the
limits of datatypes like integers, and these are stored in your C
programming enviroment in a file called limit.h.which on my system is
located at :
/usr/include/limit.h
Integer sizes aren't nearly as well defined as I would like. The
standard is a typical compromise of different corperate and academeic
intrests. In the KN King text, on page 111, the following limits are
outlined on 32 but systems
Integer data types can be defined as extended types with different sizes
as follows:
short int: -32,768 - 32768
unsigned short int: 0 - 65535
int -2,147,483.648 - 2,147,483,648
unsigned int 0 - 4,294,967,295
long int -2,147,483.648 - 2,147,483,648
unsigned long int 0 - 4,294,967,295
These numbers are byte sizes (65535 is the largest number representable
in 4 bytes).
If you have a 64 bit system, see your limits.h file for the definition
on your box.
You can drop the word "int" for longs and shorts.
long int x;
syntaxically is the same as
long x;
although we haven't yet discussed any syntax.
Computers has a special chip to work with factional numbers in decimal
notation (not an easy piece of engineering IMO). C++ has built in data
types to cope with them which have the following minimal and maximum
representations on 32 bit systems:
float 1.17 x 10^-39 - 3.40x10^38 and 6 digit precision
double 2.22 x10^-308 - 1.19x10^308 and 15 digit precision
long double very hardware specific
As I understand it, the numbers are actually stored in scientific
notation which explains a precision of 10^308 is only expressed in 15
accurate digits. These are engineering definition as defined by an
electronics standard called IEEE standard 754...more information than we
usually need. But if you need know what you limits are exactly, see the
float.h file that on my system is located at:
/usr/include/c++/4.4/tr1/float.h
and
/usr/lib/gcc/i586-suse-linux/4.4/include/float.h
There are also manual pages on your Linux system
man float.h
C++ adds one additional data type calls a Boolean type to store true or
false. In C (and in C++), processes that need to test for a true or
false view 0 as false and anything else as true. But there are issues
with that. int's are 4 bytes and signed, chars can actually be singed
and unsigned as well....so C++ adds
bool true or false
which is really like some kiind of enum operator (which hasn't been
introduced).
--
http://www.mrbrklyn.com - Interesting Stuff
http://www.nylxs.com - Leadership Development in Free Software
So many immigrant groups have swept through our town that Brooklyn, like Atlantis, reaches mythological proportions in the mind of the world - RI Safir 1998
http://fairuse.nylxs.com DRM is THEFT - We are the STAKEHOLDERS - RI Safir 2002
"Yeah - I write Free Software...so SUE ME"
"The tremendous problem we face is that we are becoming sharecroppers to our own cultural heritage -- we need the ability to participate in our own society."
"> I'm an engineer. I choose the best tool for the job, politics be damned.<
You must be a stupid engineer then, because politcs and technology have been attached at the hip since the 1st dynasty in Ancient Egypt. I guess you missed that one."
? Copyright for the Digital Millennium
|
|
