How do I completely remove a line in Rust? Not just replace it with an empty line.
In Rust, when you delete a line from a file with the following code as an example:
let mut file: File = File::open("file.txt").unwrap();
let mut buf = String::from("");
file.read_to_string(&mut buf).unwrap(); //Read the file to a buffer
let reader = BufReader::new(&file);
for (index, line) in reader.lines().enumerate() { //Loop through all the lines in the file
if line.as_ref().unwrap().contains("some text") { //If the line contains "some text", execute the block
buf = buf.replace(line.as_ref().unwrap(), ""); //Replace "some text" with nothing
}
}
file.write_all(buf.as_bytes()).unwrap(); //Write the buffer back to the file
file.txt:
random text
random text
random text
some text
random text
random text
When you run the code, file.txt turns into this:
random text
random text
random text
random text
random text
Rather than just
random text
random text
random text
random text
random text
Is there any way to completely remove the line rather than just leaving it blank? Like some sort of special character?
This part is bad-news: buf = buf.replace(line.as_ref().unwrap(), "");
This is doing a search through your entire buffer to find the line contents (without '\n') and replace it with "". To make it behave as you expect you need to add back in the newline. You can just about do this by buf.replace(line.as_ref().unwrap() + "\n", "") The problem is that lines() treats more than "\n" as a newline, it also splits on "\r\n". If you know you're always using "\n" or "\r\n" as newlines you can work around this - if not you'll need something tricker than lines().
However, there is a trickier issue. For larger files, this may end up scanning through the string and resizing it many times, giving an O(N^2) style behaviour rather than the expected O(N). Also, the entire file needs to be read into memory, which can be bad for very large files.
The simplest solution to the O(N^2) and memory issues is to do your processing line-by-line, and
then move your new file into place. It would look something like this.
//Scope to ensure that the files are closed
{
let mut file: File = File::open("file.txt").unwrap();
let mut out_file: File = File::open("file.txt.temp").unwrap();
let reader = BufReader::new(&file);
let writer = BufWriter::new(&out_file);
for (index, line) in reader.lines().enumerate() {
let line = line.as_ref().unwrap();
if !line.contains("some text") {
writeln!(writer, "{}", line);
}
}
}
fs::rename("file.txt.temp", "file.txt").unwrap();
This still does not handle cross-platform newlines correctly, for that you'd need a smarter lines iterator.
Hmm could try removing the new line char in the previous line
I have the following string array in matlab built the following way:
labels=textread(nome_tecnicas_base, '%s');
for i=1:size(labels)
temp_vector=cell(1,10);
[temp_vector{1:10}]=deal(labels{i});
final_vector=horzcat(final_vector,temp_vector);
end
I want to save this vector with each string value separated with commas (e.g., csv files) in a text file. I tried in several ways, but when I try to read it with, for example, the textread function i have the following error:
a=textread('labels-cpen-R.txt')
Error using dataread
Trouble reading number from file (row 1, field 1) ==> dct,dct,dct,dct,dct,dct,dct,dct,dct,dct,hierar
this is how my file was saved
dct,dct,dct,dct,dct,dct,dct,dct,dct,dct,hierarch-sift,hierarch-sift,hierarch-sift,hierarch-sift,hierarch-sift,hierarch-sift,hierarch-sift,hierarch sift,hierarch-sift,hierarch
sift,zernike,zernike,zernike,zernike,zernike,zernike,zernike,zernike,zernike,zernike,zernike2,zernike2,zernike2,zernike2,zernike2,zernike2,zernike2,zernike2,zernike2,zernike2,kpca,kpca,kpca,kpca,kpca,kpca,kpca,kpca,kpca,kpca,sift,sift,sift,sift,sift,sift,sift,sift,sift,sift,surf,surf,surf,surf,surf,surf,surf,surf,surf,surf,bayesianfusion0,bayesianfusion0,bayesianfusion0,bayesianfusion0,bayesianfusion0,bayesianfusion0,bayesianfusion0,bayesianfusion0,bayesianfusion0,bayesianfusion0,bks-fusion,bks-fusion,bks-fusion,bks-fusion,bks-fusion,bks-fusion,bks-fusion,bks-fusion,bks-fusion,bks-fusion,fusionvoting4,fusionvoting4,fusionvoting4,fusionvoting4,fusionvoting4,fusionvoting4,fusionvoting4,fusionvoting4,fusionvoting4,fusionvoting4,fusionvoting6,fusionvoting6,fusionvoting6,fusionvoting6,fusionvoting6,fusionvoting6,fusionvoting6,fusionvoting6,fusionvoting6,fusionvoting6,multiscale_voting,multiscale_voting,multiscale_voting,multiscale_voting,multiscale_voting,multiscale_voting,multiscale_voting,multiscale_voting,multiscale_voting,multiscale_voting,bks_rf_lvt,bks_rf_lvt,bks_rf_lvt,bks_rf_lvt,bks_rf_lvt,bks_rf_lvt,bks_rf_lvt,bks_rf_lvt,bks_rf_lvt,bks_rf_lvt,bks_svr_lvt,bks_svr_lvt,bks_svr_lvt,bks_svr_lvt,bks_svr_lvt,bks_svr_lvt,bks_svr_lvt,bks_svr_lvt,bks_svr_lvt,bks_svr_lvt,bks_svr_otsu,bks_svr_otsu,bks_svr_otsu,bks_svr_otsu,bks_svr_otsu,bks_svr_otsu,bks_svr_otsu,bks_svr_otsu,bks_svr_otsu,bks_svr_otsu,multiscale_bks_rf_lvt,multiscale_bks_rf_lvt,multiscale_bks_rf_lvt,multiscale_bks_rf_lvt,multiscale_bks_rf_lvt,multiscale_bks_rf_lvt,multiscale_bks_rf_lvt,multiscale_bks_rf_lvt,multiscale_bks_rf_lvt,multiscale_bks_rf_lvt,multiscale_bks_svr_lvt,multiscale_bks_svr_lvt,multiscale_bks_svr_lvt,multiscale_bks_svr_lvt,multiscale_bks_svr_lvt,multiscale_bks_svr_lvt,multiscale_bks_svr_lvt,multiscale_bks_svr_lvt,multiscale_bks_svr_lvt,multiscale_bks_svr_lvt
How can I save this vector and how can I read this file properly?
try textscan for reading and fprintf for writing
from the matlab documentation:
fileID = fopen('data.csv');
C = textscan(fileID,'%f %f %f %f %u8 %f',...
'Delimiter',',','EmptyValue',-Inf);
so in your case:
textscan(fileID,'%s', 'Delimiter', ',')
edit: for writing data to a file, you can use fprintf with a file identifier:
fileID= fopen('data.csv', 'w') ;
fprintf(fileID, '%s,', data{1,1:end-1}) ;
fprintf(fileID, '%s\n', data{1,end}) ;
fclose(fileID)