need your help about writing data to file

[john_tran]

I've read the LZW compression algorithm. And i know i must use 12 bit number to code the string.
But i dont know how to write this code 12 bit to file. But C++ just accepts bytes. How can i do that with bit. Someone tell me do this :
write 16bit (2 bytes) , certainly remaining 4 bit from the next code. And add the next 16 bit to file, surely contains 8 bit from the next code, and so one. I think it is right, isn it?
But how i can do that?
Please suggesting me all the technologies i have to learn for this purpose.
Or give me the link to page that contain this topic or tutorial.
Thanks

[Valmont]

The standard (C++) never defined a byte as 8 bits. It defines as:
A byte is AT LEAST 8 bits.

So if you are working with 12-bit bytes then you can still use the normal way of stream manipulation.

[joe_bruin]

A byte is AT LEAST 8 bits.

c++ does not define a byte at all.
a char must be of size 1. however many bits are in "1" is undefined, but a char must be able to store the "basic character set". this can be accomplished in 7 bits.

c99: 6.2.5

2. An object declared as type char is large enough to store any member of the basic execution character set. If a member of the required source character set enumerated in 5.2.1 is stored in a char object, its value is guaranteed to be positive. If any other character is stored in a char object, the resulting value is implementation-defined but shall be within the range of values that can be represented in that type.

as to the original question, see the bit-shift operators (<<, >>)

[Valmont]

Yes you are right. I just opened the iso standard for C++.

but a char must be able to store the "basic character set". this can be accomplished in 7 bits.

Yep, a 6-bit system does have problems has C/C++.
Thanks for the correction.

[john_tran]

I still dont know any thing.
Tell me more about that. And if you can, give me an example or practice.
Thanks all.

[Valmont]

I don't know about LZW.

Is this any help?
http://members.aol.com/rf21exe/gif.htm

[john_tran]

Oh, I knew how LZW does, but i dont know how to write it in C++, because of bits as i mentioned.
what techs i do when i break bits. Such as :
Put 12 bit of a 12bit-number (A) and concatenate 4bit of the next 12bit-number (B) into 2 bytes and write it to file, and then put 8bit remaining of 12bit-number (B) and concatenate 8bit of the next 12bit-number (C) into 2 bytes and write it to file. After then, i continue put 4bit remaining of 12bit-number (C) and concatenate 12bit of the next 12bit-number (E) into 2 bytes and write it to file. So on, just repeat the loop if i want to add more.
But How i can do that in C++, what functions or technologies i have to learn?
Thanks for reply.

[Valmont]

Originally posted by joe_bruin
c++ does not define a byte at all.
a char must be of size 1. however many bits are in "1" is undefined, but a char must be able to store the "basic character set". this can be accomplished in 7 bits.

c99: 6.2.5

2. An object declared as type char is large enough to store any member of the basic execution character set. If a member of the required source character set enumerated in 5.2.1 is stored in a char object, its value is guaranteed to be positive. If any other character is stored in a char object, the resulting value is implementation-defined but shall be within the range of values that can be represented in that type.

as to the original question, see the bit-shift operators (<<, >>)

Thinking again:
C++ is not based on C99 but C90.
I am sure that a minimal of 8 bits are guaruanteed. The value -127 +127 are guaranteed for a byte.
The trick is that CHAR_BIT = 8.
- number of bits for the smallest object that is not a bit-field (byte)
Defined here:
5.2.4.2.1 Sizes of integer types

[Valmont]

Originally posted by john_tran
Oh, I knew how LZW does, but i dont know how to write it in C++, because of bits as i mentioned.
what techs i do when i break bits. Such as :
Put 12 bit of a 12bit-number (A) and concatenate 4bit of the next 12bit-number (B) into 2 bytes and write it to file, and then put 8bit remaining of 12bit-number (B) and concatenate 8bit of the next 12bit-number (C) into 2 bytes and write it to file. After then, i continue put 4bit remaining of 12bit-number (C) and concatenate 12bit of the next 12bit-number (E) into 2 bytes and write it to file. So on, just repeat the loop if i want to add more.
But How i can do that in C++, what functions or technologies i have to learn?
Thanks for reply.

You have to learn the ways of shifting like Joe mentioned:
>> , <<
And masking might come in handy: &

I am not so good with this myself. Sorry.

[john_tran]

Yes, I will do as your Command!!!
(I think Bit Operator will solve all my problems)