Printable View
firstly, filehandles
When a file contains text, for example, "hello, person"
an ofstream being used like this: ofstream>>variable would only read "hello,"
my thinking is that a loop like this:
woudl read in the entire fileCode:while(variable!=NULL){
ofstream>>variable;
}//end while
could someone cofirm that, or give me a different way to read in the whole file?
second, how do I get the html code from a webpage?
thanx
The end of file is indicated by an EOF char, not NULL, which is a gauranteed location in memory, or a null char, which isn't the same as NULL. When the input stream finds the EOF char it will cause the stream to fail, and set the EOF bit. Therefore, the following syntax is commonly use to read in file data.
Code:while(inputStreamName >> variableName)
{
//do something with variableName other than overwriting it each time through loop
}
//determine why stream failed
if(inputStreamName.eof())
//eof() evaluates to true if EOF bit is flipped
cout << "file read in successfully" << endl;
else
cout << "error reading in file" << endl;
so...how do I concatenate the input?
would this work?:
Code:while(! inputStreamName.eof()){
inputStreamName>>temp;
variable+=temp;
}//end while
That really depends on what you are reading from the file. Concatenating strings like that requires you to use a string object supporting the += operator. If its all numerical data, then += will give you the sum, so on and so forth.Quote:
Originally Posted by pktcperlc++java
There is no easy "do it this way all the time answer". Just make sure to save off what was read during each iteration so its not overwritten in the next iteration.
PK
I should have been clearer
I meant to ask if the evaluation in the while loop was valid
Yes, it is valid. No, it will not work as you expect. Read the FAQ and do as elad suggested or your loop will go one too many times.
In addition, by using >> you are ignoring whitespace, and so when you add on the data to the variable then the whitespace is missing. There are many solutions to this, including reading the input without skipping whitespace, adding the whitespace back in if you know what it is, reading in the data from the file a character at a time, and using getline to read a line at a time.
Do you mean displaying file contents? If you do you can do this:
i think that's right...Code:ofstream fin("FILENAME");
char ch;
while(fin.get(ch))
{
cout<<ch;
}
fin.close();
so, modified, would this work?
Code:ofstream fin("FILENAME");
char ch;
while(fin.get(ch)){
storageString+=ch;
}//end while
fin.close();
are you trying to get the contents of the file written into a string? I had the same question not too long ago and nobody gave me a simple solution so i used this and it worked:
i think that's right (im not at home right now so i cant check) but it worked for me.Code:...
ifstream fin("filename");
char ch;
char newstring[100];
int x;
x=0;
while(fin.get(ch))
{
cout<<ch;
newstring[x]=ch;
x++;
}
...
would this work? maybe?
I was assuming that ifstream had a function like pos that might say what the current position in the file might beCode:#include <fstream>
#include <string>
#include <cstring>
using namespace std;
using std::string;
string readFile(string filename){
string file;
char *inread;
ifstream inf (filename);
while(inf.pos()<inf.eof()){
inf>>inread[inf.pos];
}//end while
inf.close()
for(int i=0;i<strlen(inread);i++){
file+=inread[i];
}//end for
return file;
}//end readFile
if it doesn't inf.pos() coudl be replaced with an incrementing value to mark position (starting at 0)
sry, I forgot to add whitespaceQuote:
for(int i=0;i<strlen(inread);i++){
file+=inread[i];
}//end for
after thethere should be an if statement:Code:file+=inread[i]
Code:if(i!=strlen(inread)-1){
file+=" ";
}//end if
> would this work? maybe?
Nope - inread is an uninitialised pointer.
how bout this?
does pos() actually exit?Code:#include <fstream>
#include <string>
#include <vector>
using namespace std;
using std::string;
string readFile(string filename){
string file;
vector <char> inread;
ifstream inf (filename);
while(inf.pos()<inf.eof()){
inf>>inread[inf.pos];
}//end while
inf.close()
for(int i=0;i<inread.size();i++){
file+=inread[i];
if(i!=inread.size()-1){
file+=" ";
}//end if
}//end for
return file;
}//end readFile
Why do you need to assume anything? Look at the documentation, I know it may seem like a novel idea but it seems to work for everyone else.Quote:
I was assuming that ifstream had a function like pos that might say what the current position in the file might be
[edit]
BTW this is the 3rd time you have made this post - I mean exact name and everything. Whats the deal?
[/edit]
File pointers do exist, for example fseek() in C or seekp(), seekg(), tellg() and tellp())in C++. They wouldn't seem to the most efficient way to read a whole file, but if you knew exactly how much of a file you wanted to read, or if you wanted to start reading, stop reading, etc at a particular location in a file, then using them would make more sense. Such knowledge is more readily available if the file contents are perfectly ordered.
eof() returns a non-zero value if the end of file has been reached, and a zero value if the end of file hasn't been reached. What the non-zero value is is not indicated and I wouldn't recommend it be measured against as in
while(inf.pos()<inf.eof())
I believe I gave you a rather simple solution in the end...Quote:
Originally Posted by Junior89
Only as long as your files don't contain more than 99 characters, and only as long as you don't assume that the result is a NUL-terminated string.Quote:
I had the same question not too long ago and nobody gave me a simple solution so i used this and it worked:
i think that's right (im not at home right now so i cant check) but it worked for me.
It worked for me because it was for an encryption program and the same program that read the files wrote the files with a 10,000 char limit. Your idea was to copy each of the characters into a string right? That's exactly what my method does.