How to read each lines of the input and output them in sorted order?

Question

Duplicate lines should be printed the same number of times they occur in the input. Special care needs to be taken so that a file with a lot of duplicate lines does not use more memory than what is required for the number of unique lines.

I've tried all the collection interfaces but none seems to be working for this question :( Can someone please help me?? Thanks.

The code below is memory inefficient, as it stores duplicate lines in the PriorityQueue. Hope this helps

public static void doIt(BufferedReader r, PrintWriter w) throws IOException {
    PriorityQueue<String> s=new PriorityQueue<String>();


    String   line;
    int n=0;
    while ((line = r.readLine()) != null) {


        s.add(line);
        n++;

    while (n!=0) {
        w.println(s.remove());
        n--;


    }


}

Unfortunately they didn't teach mind-reading as part of my comp sci degree. Please explain exactly what the homework problem is, show us the code you've tried so far, and where exactly you're stuck. — musical_coder, Sep 20 '13 at 05:16
Post you [SSCCE](http://sscce.org/) that shows what you have tried. The collections interface is the place to start. I would probably use an ArrayList. — camickr, Sep 20 '13 at 05:16
For this particular example, there are a few approaches that could work. I suggest keeping your code and interning each line. — chrylis -cautiouslyoptimistic-, Sep 20 '13 at 05:23
Also, have you actually confirmed that there's any problem here, or are you just noticing that there *might* be an efficiency issue? — chrylis -cautiouslyoptimistic-, Sep 20 '13 at 05:32
yeah, there is a program that runs to test efficiency. And it said the code ran out of memory — Manuel, Sep 20 '13 at 05:44

score 0 · Answer 1 · answered Sep 20 '13 at 05:18

0

You are looking for Insertion sort, which is an online algorithm, assuming lines are being inputted on the fly, if its an offline case(text file which isn't being modified on the fly), well you can use any sort algorithm, thinking of each line as a String, and the complete file as an Array of strings. Sort the array, then loop through it while printing, and then you got, sorted lines printed.

answered Sep 20 '13 at 05:18

Sinn

274
1
7

The duplicate lines need to be 'outputed' as well – Manuel Sep 20 '13 at 05:22

chrylis -cautiouslyoptimistic- · Accepted Answer · 2013-09-20T05:42:57.133

0

The ideal approach would be to use a sorted multiset, such as Guava's TreeMultiset.

If you're not allowed to use external libraries, you can replace s.add(line) with s.add(line.intern()). This tells the JVM to put a copy of each unique line into the String pool and share the same object among all the references.

Note that putting Strings into the pool may cause them to stick around for a long time, which can cause problems in long-running applications, so you don't want to do this casually in a production application, but for your homework problem it's fine. In the case of a production application, you'd want to put the Strings into a SortedMap where the value is the count of times the line appeared, but this is more complicated to code correctly.

edited Sep 20 '13 at 05:42

answered Sep 20 '13 at 05:32

chrylis -cautiouslyoptimistic-

75,269
21
115
152

what if there are more than two duplicate lines, randomly? – Manuel Sep 20 '13 at 05:42
@Manuel I don't understand your question. Were you confused by the typo in my link syntax that looked like an array? – chrylis -cautiouslyoptimistic- Sep 20 '13 at 05:43
I was told: intern() in java.lang.String cannot be applied to (java.lang.String) – Manuel Sep 20 '13 at 05:46
@Manuel Sorry, it's late. Corrected. – chrylis -cautiouslyoptimistic- Sep 20 '13 at 05:47

How to read each lines of the input and output them in sorted order?

2 Answers2