input formatted file

[Jun]

hello,

I've a data file

========
4 1 1 1 1
5 1 1 1 1 1
2 1 1
========

first column represents the how many values in this line, and then
followed by values. I've seen some codes using

*************** *************** *
ifstream inf

while(inf >data1 >data2)
*************** *************** *
to read data, just very beautiful and simple code, could anyone give
me some advices for my case? thank you in advance.
Jun

[Ron AF Greve]

Hi,

Not tested, but my guess would be something like:

#include <vector>
#include <iostream>
using namespace std;

//........
try
{
unsigned long Cnt = 0;
while( inf >Cnt )
{
vector<unsigned longData;
for( unsigned long CntData = 0; CntData < Cnt; ++CntData )
{
unsigned long Item = 0;
if( !( inf >Item ) )
{
throw "Unexpecete d end of file";
}
Data.push_back( Item );
}
// Do something with data
}
}
catch( const char *Error )
{
cerr << Error << endl;
}
Regards, Ron AF Greve

http://www.InformationSuperHighway.eu

"Jun" <ju*****@gmail. comwrote in message
news:81******** *************** ***********@y5g 2000hsf.googleg roups.com...

hello,

I've a data file

========
4 1 1 1 1
5 1 1 1 1 1
2 1 1
========

first column represents the how many values in this line, and then
followed by values. I've seen some codes using

*************** *************** *
ifstream inf

while(inf >data1 >data2)
*************** *************** *
to read data, just very beautiful and simple code, could anyone give
me some advices for my case? thank you in advance.
Jun

[Jim Langston]

"Jun" <ju*****@gmail. comwrote in message
news:81******** *************** ***********@y5g 2000hsf.googleg roups.com...

hello,

I've a data file

========
4 1 1 1 1
5 1 1 1 1 1
2 1 1
========

first column represents the how many values in this line, and then
followed by values. I've seen some codes using

*************** *************** *
ifstream inf

while(inf >data1 >data2)
*************** *************** *
to read data, just very beautiful and simple code, could anyone give
me some advices for my case? thank you in advance.

One thing I would do is use std::getline( inf, somestring );
then put it into a stringstream.
std::stringstre am Data( somestring );
At this point you can check if the number of values are correct.

int Count;
if ( Data >Count )
{
int Value;
int ValueCount;
while ( Data >Value )
{
ValueCount++;
// dosomething with Value
}
if ( ValueCount != Count )
{
// throw error here. Number at beginning does not match
// number of values.
}
}

If you don't use a stringstream but just do
while ( inf >Number )
you can't be sure if you're reading data or the count. If you were supposed
to be reading 5 values, such as the lines
5 1 1 1 1
2 1 1

It would read the next lines 2 as a value, grab the next 1 as a count, then
read the next 1 as a value, if you understand.

but getline will only read one line, which you can validate the count if you
need to.

[Jun]

On Nov 24, 5:37 pm, "Jim Langston" <tazmas...@rock etmail.comwrote :

"Jun" <junh...@gmail. comwrote in message

news:81******** *************** ***********@y5g 2000hsf.googleg roups.com...

hello,

I've a data file

========
4 1 1 1 1
5 1 1 1 1 1
2 1 1
========

first column represents the how many values in this line, and then
followed by values. I've seen some codes using

*************** *************** *
ifstream inf

while(inf >data1 >data2)
*************** *************** *
to read data, just very beautiful and simple code, could anyone give
me some advices for my case? thank you in advance.

One thing I would do is use std::getline( inf, somestring );
then put it into a stringstream.
std::stringstre am Data( somestring );
At this point you can check if the number of values are correct.

int Count;
if ( Data >Count )
{
int Value;
int ValueCount;
while ( Data >Value )
{
ValueCount++;
// dosomething with Value
}
if ( ValueCount != Count )
{
// throw error here. Number at beginning does not match
// number of values.
}

}

If you don't use a stringstream but just do
while ( inf >Number )
you can't be sure if you're reading data or the count. If you were supposed
to be reading 5 values, such as the lines
5 1 1 1 1
2 1 1

It would read the next lines 2 as a value, grab the next 1 as a count, then
read the next 1 as a value, if you understand.

but getline will only read one line, which you can validate the count if you
need to.

the following code
=============== =
std::stringstre am Data( somestring );
=============== =

just doesn't work

[Jim Langston]

"Jun" <ju*****@gmail. comwrote in message
news:02******** *************** ***********@x69 g2000hsx.google groups.com...

On Nov 24, 5:37 pm, "Jim Langston" <tazmas...@rock etmail.comwrote :

>"Jun" <junh...@gmail. comwrote in message

news:81******* *************** ************@y5 g2000hsf.google groups.com...

hello,

I've a data file

========
4 1 1 1 1
5 1 1 1 1 1
2 1 1
========

first column represents the how many values in this line, and then
followed by values. I've seen some codes using

*************** *************** *
ifstream inf

while(inf >data1 >data2)
*************** *************** *
to read data, just very beautiful and simple code, could anyone give
me some advices for my case? thank you in advance.

One thing I would do is use std::getline( inf, somestring );
then put it into a stringstream.
std::stringstr eam Data( somestring );
At this point you can check if the number of values are correct.

int Count;
if ( Data >Count )
{
int Value;
int ValueCount;
while ( Data >Value )
{
ValueCount++;
// dosomething with Value
}
if ( ValueCount != Count )
{
// throw error here. Number at beginning does not match
// number of values.
}

}

If you don't use a stringstream but just do
while ( inf >Number )
you can't be sure if you're reading data or the count. If you were
supposed
to be reading 5 values, such as the lines
5 1 1 1 1
2 1 1

It would read the next lines 2 as a value, grab the next 1 as a count,
then
read the next 1 as a value, if you understand.

but getline will only read one line, which you can validate the count if
you
need to.

the following code
=============== =
std::stringstre am Data( somestring );
=============== =

just doesn't work

std::string somestring("5 1 1 1 1");
std::stringstre am Data( somestring );

make sure somestring is a std::string
you could call it anything, "somestring " was just a foo name.

Make sure you
#include <string>
#include <sstream>

[James Kanze]

On Nov 24, 5:37 pm, "Jim Langston" <tazmas...@rock etmail.comwrote :

"Jun" <junh...@gmail. comwrote in message

news:81******** *************** ***********@y5g 2000hsf.googleg roups.com...

I've a data file

========
4 1 1 1 1
5 1 1 1 1 1
2 1 1
========

first column represents the how many values in this line, and then
followed by values. I've seen some codes using

*************** *************** *
ifstream inf

while(inf >data1 >data2)
*************** *************** *
to read data, just very beautiful and simple code, could anyone give
me some advices for my case? thank you in advance.

One thing I would do is use std::getline( inf, somestring );
then put it into a stringstream.

Why a stringstream, and not an istringstream? You don't want to
write to it?

std::stringstre am Data( somestring );
At this point you can check if the number of values are correct.

int Count;
if ( Data >Count )
{
int Value;
int ValueCount;
while ( Data >Value )
{
ValueCount++;
// dosomething with Value
}
if ( ValueCount != Count )
{
// throw error here. Number at beginning does not match
// number of values.
}
}

Even easier:

std::istringstr eam s( line ) ;
size_t count ;
s >count ;
std::vector< int data((std::istr eam_iterator< int >( s )),
(std::istream_i terator< int >()) ) ;
if ( ! s ) {
// Some input error...
} else if ( data.size() != count ) {
// syntax error
}

--
James Kanze (GABI Software) email:ja******* **@gmail.com
Conseils en informatique orientée objet/
Beratung in objektorientier ter Datenverarbeitu ng
9 place Sémard, 78210 St.-Cyr-l'École, France, [phone]

[Jim Langston]

"James Kanze" <ja*********@gm ail.comwrote in message
news:fb******** *************** ***********@j20 g2000hsi.google groups.com...
On Nov 24, 5:37 pm, "Jim Langston" <tazmas...@rock etmail.comwrote :

"Jun" <junh...@gmail. comwrote in message

news:81******** *************** ***********@y5g 2000hsf.googleg roups.com...

I've a data file

========
4 1 1 1 1
5 1 1 1 1 1
2 1 1
========

first column represents the how many values in this line, and then
followed by values. I've seen some codes using

*************** *************** *
ifstream inf

while(inf >data1 >data2)
*************** *************** *
to read data, just very beautiful and simple code, could anyone give
me some advices for my case? thank you in advance.

One thing I would do is use std::getline( inf, somestring );
then put it into a stringstream.

Why a stringstream, and not an istringstream? You don't want to
write to it?

----------

I don't use streams that often other than stringstream where I do tend to do
both input and output and the full blown stringstream is required. I'm not
100% sure of the advantages of using istringstream over stringstream though,
other than maybe some overhead costs.

std::stringstre am Data( somestring );
At this point you can check if the number of values are correct.

int Count;
if ( Data >Count )
{
int Value;
int ValueCount;
while ( Data >Value )
{
ValueCount++;
// dosomething with Value
}
if ( ValueCount != Count )
{
// throw error here. Number at beginning does not match
// number of values.
}
}

Even easier:

std::istringstr eam s( line ) ;
size_t count ;
s >count ;
std::vector< int data((std::istr eam_iterator< int >( s )),
(std::istream_i terator< int >()) ) ;
if ( ! s ) {
// Some input error...
} else if ( data.size() != count ) {
// syntax error
}

[James Kanze]

On Nov 24, 9:52 pm, "Jim Langston" <tazmas...@rock etmail.comwrote :

"James Kanze" <james.ka...@gm ail.comwrote in message

[...]

Why a stringstream, and not an istringstream? You don't want to
write to it?

I don't use streams that often other than stringstream where I
do tend to do both input and output and the full blown
stringstream is required. I'm not 100% sure of the advantages
of using istringstream over stringstream though, other than
maybe some overhead costs.

Two important benefits: it expresses your intent better, and if
you slip up, and type a >instead of a <<, the compiler will
complain. (The < and the keys are adjacent on a US keyboard,
so if your right hand is slightly out of position, it's a likely
typo.)

It's interesting that you usually use both directions, however.
I've never had a case where I've needed a bi-directional
iostream, of any type. I can imagine them with fstream,
however, at least if you open the fstream in binary mode. I
can't even image a case where you would want one for a
stringstream, however. Are you using them as some sort of
temporary file? If so, why---what's the advantage compared to
simply saving the raw data in its internal format?

--
James Kanze (GABI Software) email:ja******* **@gmail.com
Conseils en informatique orientée objet/
Beratung in objektorientier ter Datenverarbeitu ng
9 place Sémard, 78210 St.-Cyr-l'École, France, [phone]

[Jim Langston]

"James Kanze" <ja*********@gm ail.comwrote in message
news:54******** *************** ***********@n20 g2000hsh.google groups.com...
On Nov 24, 9:52 pm, "Jim Langston" <tazmas...@rock etmail.comwrote :

"James Kanze" <james.ka...@gm ail.comwrote in message

[...]

Why a stringstream, and not an istringstream? You don't want to
write to it?

I don't use streams that often other than stringstream where I
do tend to do both input and output and the full blown
stringstream is required. I'm not 100% sure of the advantages
of using istringstream over stringstream though, other than
maybe some overhead costs.

Two important benefits: it expresses your intent better, and if
you slip up, and type a >instead of a <<, the compiler will
complain. (The < and the keys are adjacent on a US keyboard,
so if your right hand is slightly out of position, it's a likely
typo.)

It's interesting that you usually use both directions, however.
I've never had a case where I've needed a bi-directional
iostream, of any type. I can imagine them with fstream,
however, at least if you open the fstream in binary mode. I
can't even image a case where you would want one for a
stringstream, however. Are you using them as some sort of
temporary file? If so, why---what's the advantage compared to
simply saving the raw data in its internal format?

-------------------------

template<typena me T, typename F T StrmConvert( const F from )
{
std::stringstre am temp;
temp << from;
T to = T();
temp >to;
return to;
}

[James Kanze]

On Nov 26, 4:56 pm, "Jim Langston" <tazmas...@rock etmail.comwrote :

"James Kanze" <james.ka...@gm ail.comwrote in message

[...]

It's interesting that you usually use both directions, however.
I've never had a case where I've needed a bi-directional
iostream, of any type. I can imagine them with fstream,
however, at least if you open the fstream in binary mode. I
can't even image a case where you would want one for a
stringstream, however. Are you using them as some sort of
temporary file? If so, why---what's the advantage compared to
simply saving the raw data in its internal format?

template<typena me T, typename F T StrmConvert( const F from )
{
std::stringstre am temp;
temp << from;
T to = T();
temp >to;
return to;
}

That's boost::lexical_ cast. Which is probably what I'd use if I
ever needed it. I'll admit that I never saw the utility of
boost::lexical_ cast either. And I'm rather sceptical of using
iostream inserters and extractors for this, too.

Consider something like:

std::complex< double z( 1.0, 0.0 ) ;
double x = boost::lexical_ cast< double >( z ) ;

The result should obviously be 1.0. Your code will return 0.0,
and boost::lexical_ cast will raise some sort of exception.

IMHO, the example showcases both problems:

-- converting a type A to a type B makes no sense in
general---you have to define exactly what is meant, for each
case (What should std::complex< double >( 0.0, 1.0 )
return when converted to double?), and

-- whatever definition you use, the >and << won't necessarily
do the trick.

(I wonder if this isn't some sort of larger anti-pattern:
providing a generic solution for a problem which doesn't have a
generic solution.)

--
James Kanze (GABI Software) email:ja******* **@gmail.com
Conseils en informatique orientée objet/
Beratung in objektorientier ter Datenverarbeitu ng
9 place Sémard, 78210 St.-Cyr-l'École, France, [phone]