The first row of my data file is the headers for the three (or four) columns of data:
%b02_a08 b02_a08_counts b02_a08_eu
As you can see, the columns are TAB delimited. When I use getline(), it mashes the whole line together. If I use getline(,,'\t'), it returns %b02_a08 for the line. I need the line separated out, because I read each header.
If anyone knows of a way to read the headers from a data file and put them into a vector<string> so that I can run Regex on them to find the one with "eu" in it, that is all I need to do. Well, that and read the columns of data, which I will probably run into the same TAB delimited problem, there.
Use the stream Luke...
as in,
// process the header
std::getline(some_input_stream, line))
{
std::istringstream str(line);
// Now reaad each field
str >> header_1 >> header_2 >> header_3;
if (str >> header_4)
{
// process the optional column
have_col_4 = true;
}
}
// Now to read the rows
while(some_input_stream >> col_1 >> col_2 >> col_3)
{
if (have_col_4)
some_input_stream >> col_4;
}
The last block can be done with std::getline too (i.e. read a line, construct a stream, then read from the stream), however this tends to be somewhat slower...
Related
I am working on a coding project where I sort and organize data from a text file, and I cannot get the getline() function to read past the first line.
The idea is to capture the entire line, split it into 3 sections, assign it to an object, then move on. I can do everything except get getline() to work properly, here is the snippet of code I am having trouble with:
ifstream fin;
fin.open("textFile.txt");
while (!fin.eof()) // while loop to grab lines until the end of file is reached
{
getline(fin, line);
fin >> first >> last >> pace; // assigning the data to their respective variables
ClassObject obj(first, last, pace); // creating an object with those variables
ClassVector.push_back(obj); // assignment object to vector
}
This has been the closest I have gotten to reading every line while also sorting the data into a vector, but as I mentioned before, getline() will read line 1, and skip the rest of the file (1000 lines).
what you can do is rather than using !fin.eof(). i prefer to use something similar to this:
ifstream file ( fileName.c_str() );
while (file >> first >> last >> pace ) // assuming the file is delimited with spaces
{
// Do whatever you want with first, last, and pace
}
The "While loop" will keep reading the next line until we reach the end of the file.
If the lengths of first, last, pace are constant, you can also just get the contents of the line (in a string variable) and use substring on it, but this only works in the specific case that the lengths are constant throughout the entire file.
I'm writing a function which reads a CSV file using getline() and converts data to the vector of vectors. To test it I've tried to read two files with the same delimiter: one imported from the internet and second exported from R datasets. The first few lines of each looks like:
File1.csv
User ID,Category 1,Category 2,Category 3,Category 4,Category 5,Category 6,Category 7,Category 8,Category 9,Category 10
User 1,0.93,1.8,2.29,0.62,0.8,2.42,3.19,2.79,1.82,2.42
User 2,1.02,2.2,2.66,0.64,1.42,3.18,3.21,2.63,1.86,2.32
User 3,1.22,0.8,0.54,0.53,0.24,1.54,3.18,2.8,1.31,2.5
User 4,0.45,1.8,0.29,0.57,0.46,1.52,3.18,2.96,1.57,2.86
File2.csv
"","Sepal.Length","Sepal.Width","Petal.Length","Petal.Width"
"1",5.1,3.5,1.4,0.2
"2",4.9,3,1.4,0.2
"3",4.7,3.2,1.3,0.2
"4",4.6,3.1,1.5,0.2
However getline() works only for the first one. In second case it simply returns white space. The function performs similar even if I copy single lines from one file to another (of course adding or removing additional colums) -- the rows from file1 will be always properly read while those from file2 never. I've even tried removing " chars, but without much improvement. However switching from comas to '\t' solves the problem.
I'm curious what's the difference between those two files that makes such different outcome?
The source code of my function:
vector<vector<string>> readData(string fileName,int firstLine,char delimeter){
//Open data file
fstream fin;
fin.open(fileName, ios::in);
//Data stored in 2d vector of strings
vector<vector<string>> data;
vector<string> row;
string line,word,temp;
//Read data
int i=0;
while(fin>>temp){
row.clear();
//Read line and store in 'line'
getline(fin,line);
//Don't read first n lines
if (i<firstLine){
i++;
continue;
}
cout<<line<<endl;
//Break words
stringstream s(line);
//Read every column and store in in 'word;
while(getline(s,word,delimeter)){
row.push_back(word);
}
//Append row to the data vector
data.push_back(row);
}
//Close file
fin.close();
return data;
}
The problem is here:
while(fin>>temp){
row.clear();
//Read line and store in 'line'
getline(fin,line);
fin >> temp reads everything till the first space or newline. It is not clear why you do that as only with getline(fin,line) you then try to read the full line and you are not using temp. In the first file fin>>temp consumes only "User", in the second file it consumes the full line, because there are no spaces.
If you look at the read data from the first file you will also notice that the first part of each line is missing.
Tip: Use more meaningful names for your variables. I didn't manage to fully understand your logic, because variables named s and the presence of row and line at the same time causes me headaces.
I currently have a text file that is as follows:
12 6 4 9
It is a very simple text file since I want to just get one line working and then maybe expand to multiple lines later. Extra aside: this is for a RPN calculator I am working on.
I want to go through this text file character by character. The way I currently have it implemented is with a simple while loop:
string line;
while (!infile.eof()){
getline(infile, line);
if (isdigit(line[0])){
rpn_stack.push_back(atof(line.c_str()));
}
}
rpn_stack is a vector since I will not be using the built in stack libraries in C++.
The problem I am currently having is that the output is just outputting "12". Why is this?
Is there a way that I can traverse through the file character by character instead of reading as a line? Is it breaking because it finds a white space (would that be considered the EOF)?
EDIT:
The code has been rewritten to be as the following:
string line;
while (!infile.eof()){
getline(infile, line);
for (int i = 0; i < line.size(); i++){
if (isdigit(line[i])){
rpn_stack.push_back(atof(line.c_str()));
}
}
}
The output is 12 5 different times, which is obviously wrong. Not only are there 4 items in the txt document, but only one of them is a 12. Can someone give some insight?
This will read as many doubles from infile as possible (i.e. until the end of file or until it comes across a token that isn't a double), separated by whitespace.
for (double d; infile >> d;)
rpn_stack.push_back(d);
If you need parse line-by-line, as #ooga says you will need a two-stage reader that looks something like this:
for (std::string line; getline(infile, line);) {
std::istringstream stream{line};
for (double d; stream >> d;)
rpn_stack.push_back(d);
}
Bonus hint: don't use .eof()
I have managed to skip the name section when reading values from a file with name and value pairs. But is there another way to skip the name section without declaring a dummy string to store the skipped data?
Example text file: http://i.stack.imgur.com/94l1w.png
void loadConfigFile()
{
ifstream file(folder + "config.txt");
while (!file.eof())
{
file >> skip;
file >> screenMode;
if (screenMode == "on")
notFullScreen = 0;
else if (screenMode == "off")
notFullScreen = 1;
file >> skip;
file >> playerXPosMS;
file >> skip;
file >> playerYPosMS;
file >> skip;
file >> playerGForce;
}
file.close();
}
You can use std::cin.ignore to ignore input up to some specified delimiter (e.g., a new-line, to skip an entire line).
static const int max_line = 65536;
std::cin.ignore(max_line, '\n');
While many people recommend specifying a maximum of something like std::numeric_limits<std::streamsize>::max(), I do not. If the user accidentally points the program at the wrong file, they shouldn't wait while it consumes an inordinate amount of data before being told something's wrong.
Two other points.
Don't use while (!file.eof()). It mostly leads to problems. For a case like this, you really want to define a struct or class to hold your related values, define an operator>> for that class, then use while (file>>player_object) ...
The way you're reading right now really tries to read a "word" at a time, not a whole line. If you want to read a whole line, you probably want to use std::getline.
I am doing a homework assignment where we are to read company data from a file and then process it for errors.
What I have so far I think will work with the first line, but I'm not sure how to make it read each line after. Each line is a record with ID, name, and payments. Basically I want to know how I can skip to the next line after I've processed the first. I haven't included the error checking yet but I think it will be in the last do while loop after 1 record is read. If the information read into each variable is wrong I can check it and output it to either the summary file or error file.
void processFile()
{
string filename;
ifstream recordFile;
ofstream summary("summary.dat");
ofstream error("error.dat");
cout << "Please enter a filename\n";
do
{
cin >> filename;
recordFile.open(filename);
if (!recordFile.is_open())
cout << "No file by that name. Please enter another filename\n";
}
while(!recordFile.is_open());
int ID = 0;
string firstName;
string lastName;
double payment1, payment2, payment3 = (0.00, 0.00, 0.00);
string numberOfRecords;
getline(recordFile, numberOfRecords);
do
{
ws(recordFile);
recordFile >> ID;
recordFile >> firstName;
recordFile >> lastName;
recordFile >> payment1;
recordFile >> payment2;
recordFile >> payment3;
}
while(!recordFile.eof());
}
*edit : I found part of my problem, I actually need to skip the first line and read on from that point. The first line in each file has useless data in it.
Use the getline function on the ifstream object
Two things. The first is if you're going to have to read multiple
records, and each record is a new line, the best solution is almost
always to read line by line, using std::getline, and then use
std::istringstream to break up the line (record) into the desired
fields. This has the advantage of keeping your input synchronized, even
in case of errors; you don't have to worry about how much to skp ahead
or ignore.
The second point is that you're checking for eof(). This is almost
always an error; sometimes, it will lead you to reading one line too
many, and in other cases, of ignoring the last line or field. If the
input is successful (and you can only check for end of file after
trying to input beyond it), the stream will behave as true in a
conditional context; if not, it will behave as false. So your loop
should be:
std::string line;
while ( std::getline( recordFile, line ) ) {
std::istringstream record( line );
record >> ID;
if ( ! record ) ...
// ...
}
And one final comment: all of the >> operators strip leading spaces,
so you don't need your call to ws. On the other hand, with the above
schema, you might want to do something like:
if ( record >> ws && record.get() != EOF ) {
// Unexpected garbage at end of line...
}
at the very end of your loop, to verify that there isn't extra text.