«

Hello group,

I am reading an 'mpeg file' from the socket. In read socket I specify
a 'char* buffer' to read the file. However, the content of actual data
contain '\0' characters at various places.

When I read the full content I copy the char* buffer into a string type
variable. The code looks
like

int len;
char * buf[256];
int size = 256;
string content;

len = read(sockFd, buf, size);

content = buf;

Here if I print buf anything after the '\0' is not printed.

When I copy the buf to content anything after '\0' wipes out and is not
copied. I want full
buffer to be copied to content and later on down the line I want to
save it to a file including
any '\0' characters.

I also want to print the full buffer on the stdout including anything
after '\0' character.

Folks! any suggestion. This bug is stopping my program to proceed.

Thx.

arun

* nagrik:
> Hello group,
>
> I am reading an 'mpeg file' from the socket. In read socket I specify
> a 'char* buffer' to read the file. However, the content of actual data
> contain '\0' characters at various places.
>
> When I read the full content I copy the char* buffer into a string type
> variable. The code looks
> like
>
> int len;
> char * buf[256];
> int size = 256;
> string content;
>
> len = read(sockFd, buf, size);
>
> content = buf;

string content const( buf, buf + len );


--
A: Because it messes up the order in which people normally read text.
Q: Why is it such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?

* Alf P. Steinbach:
> * nagrik:
>> Hello group,
>>
>> I am reading an 'mpeg file' from the socket. In read socket I specify
>> a 'char* buffer' to read the file. However, the content of actual data
>> contain '\0' characters at various places.
>>
>> When I read the full content I copy the char* buffer into a string type
>> variable. The code looks
>> like
>>
>> int len;
>> char * buf[256];
>> int size = 256;
>> string content;
>>
>> len = read(sockFd, buf, size);
>>
>> content = buf;
>
> string content const( buf, buf + len );

Sorry, transposition errors are usually at the letter level, but somehow
here at the word level.

Should be

string const content( buf, buf + len );

--
A: Because it messes up the order in which people normally read text.
Q: Why is it such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?

nagrik wrote:
> Hello group,
>
> I am reading an 'mpeg file' from the socket. In read socket I specify
> a 'char* buffer' to read the file. However, the content of actual data
> contain '\0' characters at various places.
>
> When I read the full content I copy the char* buffer into a string type
> variable. The code looks
> like
>
> int len;
> char * buf[256];
> int size = 256;
> string content;
>
> len = read(sockFd, buf, size);
>
> content = buf;
>
> Here if I print buf anything after the '\0' is not printed.
>
> When I copy the buf to content anything after '\0' wipes out and is not
> copied. I want full
> buffer to be copied to content and later on down the line I want to
> save it to a file including
> any '\0' characters.
>
> I also want to print the full buffer on the stdout including anything
> after '\0' character.
>
> Folks! any suggestion. This bug is stopping my program to proceed.
>

You don't actually treat the data as a char* type string. The "char*"
is legacy from BSD. Treat it like a void*.

Alf P. Steinbach wrote:
> * Alf P. Steinbach:
> > * nagrik:
> >> Hello group,
> >>
> >> I am reading an 'mpeg file' from the socket. In read socket I specify
> >> a 'char* buffer' to read the file. However, the content of actual data
> >> contain '\0' characters at various places.
> >>
> >> When I read the full content I copy the char* buffer into a string type
> >> variable. The code looks
> >> like
> >>
> >> int len;
> >> char * buf[256];
> >> int size = 256;
> >> string content;
> >>
> >> len = read(sockFd, buf, size);
> >>
> >> content = buf;
> >
> > string content const( buf, buf + len );
>
> Sorry, transposition errors are usually at the letter level, but somehow
> here at the word level.
>
> Should be
>
> string const content( buf, buf + len );
>

Alf,

My content variable is already constructed so after I have buf with me,
the only operation
I am allowed is assignment.

something like

content += buf.

'buf' contains a '\0' character and there will be many more 'buf's I
would like to append to
content. Like

while ( there are more bufs available)
content += buf;

With my implementation, the assignment only copy till the end of '\0'
and later appends
append only after '\0' character and not after the whole buffer, which
I desire. Remember
the '\0' character can occur anywhere in buf and in any buf. I want
content to have full
range of values, including '\0' s.

Too bad C++ does not have a byte type. Any thoughts.

nagrik

> --
> A: Because it messes up the order in which people normally read text.
> Q: Why is it such a bad thing?
> A: Top-posting.
> Q: What is the most annoying thing on usenet and in e-mail?

* nagrik:
> Alf P. Steinbach wrote:
>> * Alf P. Steinbach:
>>> * nagrik:
>>>>
>>>> I am reading an 'mpeg file' from the socket. In read socket I specify
>>>> a 'char* buffer' to read the file. However, the content of actual data
>>>> contain '\0' characters at various places.
>>>>
>>>> When I read the full content I copy the char* buffer into a string type
>>>> variable. The code looks
>>>> like
>>>>
>>>> int len;
>>>> char * buf[256];
>>>> int size = 256;
>>>> string content;
>>>>
>>>> len = read(sockFd, buf, size);
>>>>
>>>> content = buf;
>>> string content const( buf, buf + len );
>> Sorry, transposition errors are usually at the letter level, but somehow
>> here at the word level.
>>
>> Should be
>>
>> string const content( buf, buf + len );
>
> My content variable is already constructed so after I have buf with me,
> the only operation
> I am allowed is assignment.

Why don't you just move the declaration?

Or invent a new name?

Or if neither of these options feel right, use the named 'assign' member
function?

--
A: Because it messes up the order in which people normally read text.
Q: Why is it such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?

In message <j8dwg.133411$ >, red
floyd <> writes
>nagrik wrote:
>> Hello group,
>> I am reading an 'mpeg file' from the socket. In read socket I
>>specify
>> a 'char* buffer' to read the file. However, the content of actual data
>> contain '\0' characters at various places.
>> When I read the full content I copy the char* buffer into a string
>>type
>> variable. The code looks
>> like
>> int len;
>> char * buf[256];
>> int size = 256;
>> string content;
>> len = read(sockFd, buf, size);
>> content = buf;
>> Here if I print buf anything after the '\0' is not printed.
>> When I copy the buf to content anything after '\0' wipes out and is
>>not
>> copied. I want full
>> buffer to be copied to content and later on down the line I want to
>> save it to a file including
>> any '\0' characters.
>> I also want to print the full buffer on the stdout including
>>anything
>> after '\0' character.
>> Folks! any suggestion. This bug is stopping my program to proceed.
>>
>
>You don't actually treat the data as a char* type string. The "char*"
>is legacy from BSD. Treat it like a void*.
>
He's copying it into a std::string. That takes a char*.

The real problem is that std::string has an assignment operator which
takes a const char * argument, but it stops copying at the first '\0'.
It also has a constructor which takes the same argument and has the same
problem.

The solution is that it also has constructors which take _two_
arguments, either two input iterators or a pointer and a length. Either
of these will copy the appropriate sequence without stopping at '\0':

string content(buf, size); // pointer and count
or
string content(buf, buf+size); // two input iterators

--
Richard Herring