用Java将长型（原始类型）写入文件

Question

Hi all,

I have an array of long that I would like to write into a .txt file that I can later open in gedit (one number per line). I get those values by using a subtraction of two instances of System.currentTimeMillis().

I use the following code:

BufferedWriter out = new BufferedWriter(new FileWriter("latency.txt"));
for (int i = 0; i < USER_LIMIT; ++i) {
    out.write(latency[i] + "\n");
}
out.close();

When looking at the file, I do see:

0
1
1
0

I believe the string concatenation converted the long into an integer. If I use the DataOutputStream, then I cannot read it back with gedit or any notepad/text editor, it just looks like garbage (I believe it's writing bytes).

Would anyone please let me know how I can fix my problem please?

Thank you very much!

Answer 1

There is nothing wrong with your code. What you think is in latency ... isn't.

public static void main(String[] args) throws IOException
{
    long[] latency = { 123456789000L, 234567890000L, 345678901000L };

    BufferedWriter out = new BufferedWriter(new FileWriter("latency.txt"));
    for (int i = 0; i < 3; ++i) {
        out.write(latency[i] + "\n");
    }
    out.close();
}

produces:

$ more latency.txt 
123456789000
234567890000
345678901000

When you're having a problem with code like this, it's often beneficial to write a small test case to narrow down the problem.

Answer 2

Cast to a Long and use toString:

out.write(((Long)latency[i]).toString() + "\n");

Or just use the static toString:

out.write(Long.toString(latency[i]) + "\n");

Answer 3

When you use DataOutputStream to write a long it is using a different encoding of your data than gedit understands.

One of the most common encodings used for text files is ASCII. With this each byte represents a character in a table. This table has 128 characters. When you write a string to a file using Java the way you're doing it this is what is happening and gedit understands that encoding. If you convert a long to a string the maximum size a long can be would look like: 9223372036854775807. That would take up 19 bytes.

A long in Java is 64 bits or 8 bytes. When you use DataOutputStream a long gets written to the file as 8 bytes. A text editor like gedit does not understand this encoding.

What further complicates things is encodings such as UTF-8.

Answer 4

You don't need to append the newline to the number, so you can break them apart and avoid this problem. Try this:

for ( int i = 0; i < USER_LIMIT; ++i ) {
    out.write( String.valueOf(latency[i]) ); // Or Long.toString(long)
    out.write( '\n' ); // Or simply out.newLine() 
}