Selective retrieval of numbers in a text file

[Shikha Upadhyaya]

I have a text file as follows

             total       used       free     shared    buffers     cached
Mem:          7923       7875         47          0        564       3237
-/+ buffers/cache:       4074       3848
Swap:        12287       1067      11220
Total:       20211       8943      11267

I have to retrieve only those values which fall in the mem, swap and total under total, used and free i.e. I have to ignore 0, 564, 3237, 4074, 3848. How do I do it?

I have written the following code to display the all the numbers.

package my.scan;  
import java.util.*; //importing some java classes  
import java.lang.*;  
import java.io.*;  
  
class ReadFile {
	void getFile () {
		int i = 0;
		
	       try {  
	               Scanner myFile = new Scanner(new FileReader("C:/Documents and Settings/shikhau/workspace/SanNumber/freelast.txt"));  
	               while (myFile.hasNext()) {  
	                   if (myFile.hasNextInt()) {
	                	   i = myFile.nextInt();
	                	   System.out.println(i);
	                   }
	                   else {
	                   		myFile.next();
	                   }              
	               }  
	       }  
	       catch (FileNotFoundException e){  
	           System.out.println("Sorry! This file is not found");  
	       }
	}
}

public class ScanNum {  
    public static void main(String[] args) {  
        ReadFile rf = new ReadFile();
    	rf.getFile();
          
    }  
  
}  

[Swastik Dey]

What is the field separator between the columns?

[John Jai]

1. You might want to read the text files line by line. Now you read it token by token using the next() method.
2. You have to check for matches in the lines read such as whether its for memory / swap / total.
3. Read only the first three integer tokens from the line being processed - but make sure always total used and free sizes are always present. Else you might end up reading shared, buffers, and even cached values.

Below piece of code might help you to build a scanner using the lines being read.

Scanner fileScanner = new Scanner(new FileReader("C:/test.txt"));		
Scanner lineScanner;
String line;
while (fileScanner.hasNextLine()){
	line = fileScanner.nextLine();				
	if (line.contains("Mem:")) {
		System.out.println("Memory Line -> " + line);
		lineScanner = new Scanner(line);					
	}
}

And make the lineScanner act as the myFile scanner you currently have. The lineScanner should act on all the tokens present in a single line.

[Mohamed Sanaulla]

Just an extra piece of information:
Do you have a control over the format of the file? Or is it always a text file.
If you have control over the file contents or the format- you might want to consider using xml/json format.

[Shikha Upadhyaya]

Swastik Dey wrote:What is the field separator between the columns?

A couple of space.

John Jai wrote:
1. You might want to read the text files line by line. Now you read it token by token using the next() method.
2. You have to check for matches in the lines read such as whether its for memory / swap / total.
3. Read only the first three integer tokens from the line being processed - but make sure always total used and free sizes are always present. Else you might end up reading shared, buffers, and even cached values.

Thanks!! I will work on this and get back.

Mohamed Sanaulla wrote:Do you have a control over the format of the file? Or is it always a text file.
If you have control over the file contents or the format- you might want to consider using xml/json format.

I do have control over the file format. How is json going to help? I am very new to java.

[Shikha Upadhyaya]

Mohamed Sanaulla wrote:Just an extra piece of information:
Do you have a control over the format of the file? Or is it always a text file.
If you have control over the file contents or the format- you might want to consider using xml/json format.

As an extension to what I am doing currently, I have to be able to plot this into a graph (a pie chart or a bar graph). In that case, would json be more helpful than a textfile?

[Mohamed Sanaulla]

The issue with the text file is that you may not be sure of the structure- there may be 4 spaces, or a tab or a new line seperating the data.

But with a structured document like XML or JSON you are for sure about the structure of the document and you can use parser api's already written to parse your XML or JSON documents.

[rastogi payam]

This may help you...

package com.example;

import java.io.*;  
  
class ReadFile 
{
	public static void main(String... args)
	{
		try
		{
			File f  = new File("D:\\SelectiveRetreival\\freelast.txt");
			BufferedReader br = new BufferedReader(new FileReader(f));
			String str;
			while((str=br.readLine())!=null)
			{
				if(str.contains("Mem:")||str.contains("Swap:")||str.contains("Total:"))
				{
					String[] arrStr = str.split("\\s+");
					System.out.print(arrStr[1]+" ");
					System.out.print(arrStr[2]+" ");
					System.out.print(arrStr[3]+" ");
					System.out.println();
				}
			}
		}
		catch(Exception e)
		{
			System.out.println("Error!!!");
		}
	}
}

[Shikha Upadhyaya]

Got it! Thanks