Obtain padded string containing number in hex form

[Eric Lilja]

Hello, I have an unsigned long that I need to convert to a std::string.
The unsigned long holds 32-bit checksums and sometimes the most
significant byte is 0 and in those cases the string should be zero
padded (it should always contain 8 chars) and it should display its
value in hex form without 0x in the beginning. So if the unsigned long
holds the value 0xa87d7d4 the string should contain "0a87d7d4".
I tried a combination of iomanip and stringstreams but this code
snippet still yields a string containing a decimal number (and I'm not
sure my 0-padding is working either):
std::string s;
std::stringstream ss;
ss << crc32; // crc32 is of type unsigned long
ss >> std::setw(8) >> std::hex >> s;

I guess I could read the checksum four bits at a time converting each
bit group into the corresponding hexadecimal number and gradually build
the string, but I wanted to ask you experts if there was a neater way
involving the standard library first.

/ E

[Eric Lilja]

Eric Lilja wrote:

Hello, I have an unsigned long that I need to convert to a std::string.
The unsigned long holds 32-bit checksums and sometimes the most
significant byte is 0 and in those cases the string should be zero
padded (it should always contain 8 chars) and it should display its
value in hex form without 0x in the beginning. So if the unsigned long
holds the value 0xa87d7d4 the string should contain "0a87d7d4".
I tried a combination of iomanip and stringstreams but this code
snippet still yields a string containing a decimal number (and I'm not
sure my 0-padding is working either):
std::string s;
std::stringstream ss;
ss << crc32; // crc32 is of type unsigned long
ss >> std::setw(8) >> std::hex >> s;

I guess I could read the checksum four bits at a time converting each
bit group into the corresponding hexadecimal number and gradually build
the string, but I wanted to ask you experts if there was a neater way
involving the standard library first.

I solved it using sprintf. This program shows how:
#include <cstdio>
#include <iostream>

int
main()
{
unsigned long l = 0x0a87d7d4;
char buffer[9];

// %[flags][width][.precision][modifiers]type
// type = lx = Unsigned long hexadecimal integer
// flags =none
// width = 08

std::sprintf(buffer, "%08lx", l);

std::cout << buffer << std::endl;

return 0;
}

Output:
$ ./conv.exe
0a87d7d4

But I still would like to know how to solve it using streams too, for
learning purposes.

/ E

[Jerry Coffin]

In article <11**********************@b68g2000cwa.googlegroups .com>,
mi********@gmail.com says...

Hello, I have an unsigned long that I need to convert to a std::string.
The unsigned long holds 32-bit checksums and sometimes the most
significant byte is 0 and in those cases the string should be zero
padded (it should always contain 8 chars) and it should display its
value in hex form without 0x in the beginning. So if the unsigned long
holds the value 0xa87d7d4 the string should contain "0a87d7d4".
I tried a combination of iomanip and stringstreams but this code
snippet still yields a string containing a decimal number (and I'm not
sure my 0-padding is working either):

#include <iomanip>
#include <sstream>
#include <iostream>

int main() {

unsigned crc32 = 0xa87d7d4;
std::stringstream ss;
std::string s;

ss << std::hex << std::setfill('0')
<< std::setw(8) << std::setprecision(8)
<< crc32;

// _obviously_ better than '%8.8x'

s = ss.str();
std::cout << s << std::endl;
return 0;
}

--
Later,
Jerry.

The universe is a figment of its own imagination.

[Eric Lilja]

Jerry Coffin wrote:

In article <11**********************@b68g2000cwa.googlegroups .com>,
mi********@gmail.com says...

Hello, I have an unsigned long that I need to convert to a std::string.
The unsigned long holds 32-bit checksums and sometimes the most
significant byte is 0 and in those cases the string should be zero
padded (it should always contain 8 chars) and it should display its
value in hex form without 0x in the beginning. So if the unsigned long
holds the value 0xa87d7d4 the string should contain "0a87d7d4".
I tried a combination of iomanip and stringstreams but this code
snippet still yields a string containing a decimal number (and I'm not
sure my 0-padding is working either):

#include <iomanip>
#include <sstream>
#include <iostream>

int main() {

unsigned crc32 = 0xa87d7d4;
std::stringstream ss;
std::string s;

ss << std::hex << std::setfill('0')
<< std::setw(8) << std::setprecision(8)
<< crc32;

// _obviously_ better than '%8.8x'

s = ss.str();
std::cout << s << std::endl;
return 0;
}

Seems to work just fine, thanks Jerry! Now I know how to do it using
both C++ streams and C style.

/ E

[ax]

On Sat, 24 Jun 2006 09:30:40 -0600, Jerry Coffin <jc*****@taeus.com>
wrote:

#include <iomanip>
#include <sstream>
#include <iostream>

int main() {

unsigned crc32 = 0xa87d7d4;
std::stringstream ss;
std::string s;

ss << std::hex << std::setfill('0')
<< std::setw(8) << std::setprecision(8)
<< crc32;
don't will be better:?

cout << (ss << fmt("0x%8x", '0') << crc32);

// _obviously_ better than '%8.8x'

s = ss.str();
std::cout << s << std::endl;
return 0;
}

[Jerry Coffin]

In article <4k********************************@4ax.com>, [e-mail]
says...

On Sat, 24 Jun 2006 09:30:40 -0600, Jerry Coffin <jc*****@taeus.com>

[ ... ]

ss << std::hex << std::setfill('0')
<< std::setw(8) << std::setprecision(8)
<< crc32;

don't will be better:?

cout << (ss << fmt("0x%8x", '0') << crc32);

If you have the 'fmt' function available, and it does what it looks
like it would above, it may be quite useful. But be aware that it's
not part of the standard library, nor even part of TR1, so it's not
what you'd usually think of as particularly portable.

--
Later,
Jerry.

The universe is a figment of its own imagination.

[Duane Hebert]

If you have the 'fmt' function available, and it does what it looks
like it would above, it may be quite useful. But be aware that it's
not part of the standard library, nor even part of TR1, so it's not
what you'd usually think of as particularly portable.

Is boost::format part of TR1? At any rate, it may be a useful
alternative here.

[Jerry Coffin]

In article <%f****************@weber.videotron.net>, sp**@flarn2.com
says...

[ ... ]

Is boost::format part of TR1?
I'd have to re-check to be absolutely sure, but I don't think so.
At any rate, it may be a useful alternative here.

Quite true -- there are quite a few alternatives, but I don't believe
any of them has been standardized.

--
Later,
Jerry.

The universe is a figment of its own imagination.