Reputation: 23
I am reading lines of a text file into a variable, myline and then I am trying to tokenize these lines using istringstream. However, it seems that I am losing random characters from original text file string.
cout<< myline << buff << flush; //print original text file line
istringstream iss(myline);
string sub;
while (iss >> sub) {
cout << "[" << sub << "]" << endl;
}
If you look at my output, you can see that I have the correct string from the text file, but when I use istringstream and then print the individual tokens (seen within [] brackets), some of the tokens are prematurely truncated.
#include <iostream>
[#include]
[<iostream]
#include <sstream>
[#include]
[<sstream>]
using namespace std;
[using]
[namespace]
[st]
int main()
[int]
[main(]
{
string str(" SOME LONG STRING\twith\nSPACES ");
[string]
[str("]
[SOME]
[LONG]
[STRING\twith\nSPACES]
istringstream iss(str);
[istringstream]
[iss(str);]
string s;
[strin]
while (iss >> s) {
[while]
[(iss]
[>>]
cout << "[" << s << "]" << endl;
[cout]
[<<]
["["]
[<<]
[s]
[<<]
["]"]
[<<]
[e]
}
return 0;
[retur]
}
Does anybody have any ideas what I'm doing wrong? Thanks in advance!
EDIT: Here's a version of the code that will compile fully. You can run it with any text file
#include <cstring>
#include <cstdio>
#include <iostream>
#include <iomanip>
#include <string>
#include <sstream>
#include <vector>
#include <algorithm>
#include <iterator>
using namespace std;
class MyFileReader {
public:
//constructor
MyFileReader(const char* p);
//destructor
~MyFileReader();
//getLine()
int getLine(char *buffer, int size);
//getCurrentLineNumber()
int getCurrentLineNumber();
void tokenizeLine(vector<string>& vec);
FILE * pFile;
};
//constructor
MyFileReader::MyFileReader(const char* p) {
pFile = fopen(p, "r");
}
//destructor
MyFileReader::~MyFileReader() {
fclose(pFile);
}
//getLine()
int MyFileReader::getLine(char *buffer, int size){
char *out = fgets(buffer, size, pFile);
if (out==NULL) {
return -1;
}
char *pch = strpbrk(out,"\n");
if (pch != NULL) {
return 1;
}
else {
return 0;
}
}
int MyFileReader::getCurrentLineNumber() {
static int mynumber=2;
return mynumber++;
}
//tokenizeLine
void MyFileReader::tokenizeLine(vector<string>& vec) {
string myline("");
char buff[10];
while (1) {
int result = getLine(buff, sizeof(buff));
if (result == -1 ) {
if (myline.length() > 0)
cout << myline << flush;
break;
}
else if (result == 0) {
myline += buff;
}
else if (result == 1) {
cout<< myline << buff << flush;
istringstream iss(myline);
string sub;
while (iss >> sub) {
cout << "[" << sub << "]" << endl;
}
myline = "";
}
else {
printf("PANIC");
}
}
return;
}
int main(int argc, char **argv) {
vector<string> v;
const char *filename = argv[1];
MyFileReader f(filename);
f.tokenizeLine(v);
return 0;
}
To generate the output above, I ran it on:
#include <iostream>
#include <sstream>
using namespace std;
int main()
{
string str(" SOME LONG STRING\twith\nSPACES ");
istringstream iss(str);
string s;
while (iss >> s) {
cout << "[" << s << "]" << endl;
}
return 0;
}
Upvotes: 2
Views: 372
Reputation: 2985
You line buffer is only 10 bytes long. That's not long enough to fit a full line.
void MyFileReader::tokenizeLine(vector<string>& vec) {
string myline("");
char buff[10];// this is too short
....
EDIT
As Dyp correctly points out your appending logic is not correct when you detect a \n in the input file.
Upvotes: 1
Reputation: 39111
The mistake is here:
else if (result == 1) {
cout<< myline << buff << flush;
istringstream iss(myline);
string sub;
while (iss >> sub) {
cout << "[" << sub << "]" << endl;
}
myline = "";
}
If result == 1
, that means that buff
contains \n
, it doesn't mean it only contains \n
. I.e. you drop the buffer if it contains \n
. If therefore the line happens to have n*10 (sizeof buffer) characters, your code works, else, the last characters of a line are not copied to myline
but dropped.
The quick fix would be to:
else if (result == 1) {
myline += buff; // copy the rest of the line into `myline`
cout<< myline << flush; // buff now is part of myline
istringstream iss(myline);
string sub;
while (iss >> sub) {
cout << "[" << sub << "]" << endl;
}
myline = "";
}
Though you might want to consider dropping the \n
from the buffer, like:
int MyFileReader::getLine(char *buffer, int size){
char *out = fgets(buffer, size, pFile);
if (out==NULL) {
return -1;
}
//char *pch = strpbrk(out,"\n");
char *pch = strchr(out,'\n'); // no need to search for a string
if (pch != NULL) {
*pch = '\0'; // drop the '\n'
return 1;
}
else {
return 0;
}
}
You'll have to change cout<< myline << flush;
to cout<< myline << endl;
, though.
Besides this mistake, please consider using ifstream
:
#include <iostream>
#include <fstream>
#include <sstream>
using namespace std;
int main()
{
ifstream file("test.txt");
if(!file)
{
/* error */
}else
{
string line;
while(getline(file, line))
{
istringstream iss(line);
string s;
while (iss >> s) {
cout << "[" << s << "]" << endl;
}
}
}
}
Upvotes: 1