Insert tab

[Amira]

Hi all,

I am reading from a text file and then write into an array line by
line. For that reason a writeline function has been created. When the
float numbers are written into the array I would like them to be
separated by a tab. Any idea how it might work? Thanks in advance.

Cheers,

Paul

[Victor Bazarov]

Amira wrote:

I am reading from a text file and then write into an array line by
line. For that reason a writeline function has been created.
Care to share it with us?
When the
float numbers are written into the array I would like them to be
separated by a tab. Any idea how it might work? Thanks in advance.

Yes, it will work just fine. When tab characters are inserted into the
stream between floating point numbers, the numbers will be separated by
those tabs.

V

[Amira]

--> here is the writenextvalue function ...

void writenextvalue( FILE* fh, double value, bool newline )
{
char* text = new char(20);
sprintf(text, " %lg ", value);

for (int i=0; i<strlen(text); ++i)
fputc(text[i], fh);
if (newline)
fputc(10,fh);
else
fputc(32,fh);
}

--> here the information is transfered and put into the array

for (int x=0 ; x<32 ;x++)
{
for (int z=0 ; z<16 ;z++)
{
for (int kx=0 ; kx<32 ;kx++)
{
for (int kz=0 ;kz<32 ;kz++)
{
realp[x][z]+=realpk[kx][kz]*cos(PI*2/32*(kx*x+kz*z));
imp[x][z]+=impk[kx][kz]*sin(PI*2/32*(kx*x+kz*z));

}

}

result[x][z]=sqrt(realp[x][z]*realp[x][z]+imp[x][z]*imp[x][z]);
writenextvalue( fh, result[x][z], false );
putc(\t);

}
writenewline(fh);
}

[Victor Bazarov]

Amira wrote:

--> here is the writenextvalue function ...

void writenextvalue( FILE* fh, double value, bool newline )
{
char* text = new char(20);
sprintf(text, " %lg ", value);

for (int i=0; i<strlen(text); ++i)
fputc(text[i], fh);
if (newline)
fputc(10,fh);
else
fputc(32,fh);
What do those 'fputc' do? Why it's 10 and why it's 32? What do they
mean? Why don't you simply use 'fprintf'?
} [...]

V

[Amira]

because it is writing into a stream

[Victor Bazarov]

Amira wrote:

because it is writing into a stream

I am sorry, which question are you replying to?

[Marcus Kwok]

Amira wrote:

--> here is the writenextvalue function ...

void writenextvalue( FILE* fh, double value, bool newline )
{
char* text = new char(20);
sprintf(text, " %lg ", value);

for (int i=0; i<strlen(text); ++i)
fputc(text[i], fh);
if (newline)
fputc(10,fh);
else
fputc(32,fh);

Victor Bazarov wrote: What do those 'fputc' do? Why it's 10 and why it's 32? What do they
mean? Why don't you simply use 'fprintf'?

Amira <pa********@gmail.com> wrote: because it is writing into a stream

Amira,
Please quote properly when replying to posts.

Victor was asking why you are using such things as "10" and "32" instead
of symbolic names, like '\n' for newline and '\t' for tab.

Also, since fh is a FILE*, you should be able to use fprintf, like:

(not tested)
fprintf(fh, " %lg ", value);
if (newline)
fprintf(fh, "\n");
else
fprintf(fh, "\t");
Though personally I would prefer a C++ solution:

(also not tested)
void writenextvalue(std::ofstream& fh, double value, bool newline)
{
fh /* << formatting manipulators */ << value << (newline ? '\n' : '\t');
}

--
Marcus Kwok

[John Harrison]

Amira wrote:

--> here is the writenextvalue function ...

void writenextvalue( FILE* fh, double value, bool newline )
{
char* text = new char(20);
This is wrong, it allocates a *single* character which gets the value
20. You need square brackets. But really why don't you do it like this

char text[20];

Much simpler and it stops you having a memory leak later on.

sprintf(text, " %lg ", value);

for (int i=0; i<strlen(text); ++i)
fputc(text[i], fh);
if (newline)
fputc(10,fh);
else
fputc(32,fh);
}

john

[Karl Heinz Buchegger]

Amira wrote:

A lot of good adivce given already, but
noone answered your first hand question

result[x][z]=sqrt(realp[x][z]*realp[x][z]+imp[x][z]*imp[x][z]);
writenextvalue( fh, result[x][z], false );
putc(\t);

\t is a character like any other character ('a', 'b', 'c', ... '0', '1', '\n', ... )
thus you need to enclose it in quotes

putc( '\t' );

But all of this looks mighty complicated for the simple task you are working on. Especially
having fucnctions that ... well, ... basically just call 1 function from the standard library.
Why not drop those and simply do:

for (int x=0 ; x<32 ;x++)
{
for (int z=0 ; z<16 ;z++)
{
for (int kx=0 ; kx<32 ;kx++)
{
for (int kz=0 ;kz<32 ;kz++)
{
realp[x][z]+=realpk[kx][kz]*cos(PI*2/32*(kx*x+kz*z));
imp[x][z]+=impk[kx][kz]*sin(PI*2/32*(kx*x+kz*z));
}
}
result[x][z]=sqrt(realp[x][z]*realp[x][z]+imp[x][z]*imp[x][z]);

fprintf( fh, " %lg\t", result[x][z] );
}
fprintf( "\n" );
}

--
Karl Heinz Buchegger
kb******@gascad.at