2
\$\begingroup\$

I wrote an answer for the following question and wondering if there is any better approach to it.

Using the Java language, have the function RunLength(str) take the str parameter being passed and return a compressed version of the string using the Run-length encoding algorithm. This algorithm works by taking the occurrence of each repeating character and outputting that number along with a single character of the repeating sequence.

For example: "wwwggopp" would return 3w2g1o2p. The string will not contain any numbers, punctuation, or symbols.

Code

 public class App {
    void runLength(String str) {
        HashMap<Character, Integer> hash = new HashMap<Character, Integer>();
        Character c;
        for (int i = 0; i < str.length(); i++) {
            c = str.toLowerCase().charAt(i);
            if (hash.containsKey(c)) {
                int value = hash.get(c);
                value++;
                hash.put(c, value);
            } else {
                hash.put(c, 1);
            }
        }
        int value = 0;

        StringBuffer buf = new StringBuffer();
        String temp;
        for (int j = 0; j < str.length(); j++) {
            c = str.toLowerCase().charAt(j);
            value = hash.get(c);
            temp = str.substring(j, j + 1);
            if (buf.indexOf(c.toString()) == -1) {
                buf.append(value);
                buf.append(temp);
            }

        }
        System.err.println(buf.toString());

    }

    public static void main(String[] args) {
        new App().runLength("wwwggopp");
    }
}

Output

3w2g1o2p
\$\endgroup\$
1
  • 1
    \$\begingroup\$ What's the expected output for wwwgggooopppwww? It appears that your code will not be able to handle this case. \$\endgroup\$
    – wei2912
    Oct 12, 2014 at 10:31

2 Answers 2

4
\$\begingroup\$

Avoid printing in methods, unless the purpose of the method is speicifically to print. The main purpose of the runLength method is certainly not to print. It should return the compressed value instead.

The description doesn't say that you should treat upper and lowercase characters equal. It's not correct to do that. And if you really wanted to do that, it would have been simpler to lowercase the entire input string once. This is a very wasteful operation:

for (int i = 0; i < str.length(); i++) {
    c = str.toLowerCase().charAt(i);

This would have been much better:

String lowered = str.toLowerCase();
for (int i = 0; i < lowered.length(); i++) {
    c = lowered.charAt(i);

Declare variables with the interface type, when possible. Instead of:

HashMap<Character, Integer> hash = new HashMap<Character, Integer>();

This would have been better:

Map<Character, Integer> hash = new HashMap<Character, Integer>();

Btw are you still on Java6? You should definitely migrate to at least Java7, where the above declaration becomes simply:

Map<Character, Integer> hash = new HashMap<>();

Note that in most use cases, StringBuilder is preferred over StringBuffer. StringBuffer is synchronized, StringBuilder is not, which makes it faster. In this program you don't need synchronization when building the compressed string.

Suggested implementation

This is simpler, without a hashmap:

public String compress(String str) {
    if (str.isEmpty()) {
        return "";
    }

    char[] chars = str.toCharArray();
    StringBuilder builder = new StringBuilder();

    int count = 1;
    char prev = chars[0];
    for (int i = 1; i < chars.length; i++) {
        char current = chars[i];
        if (current == prev) {
            count++;
        } else {
            builder.append(count).append(prev);
            count = 1;
        }
        prev = current;
    }
    return builder.append(count).append(prev).toString();
}

Unit testing

It's always good to have unit tests to verify correctness:

@Test
public void test_aabcccccaaa() {
    assertEquals("2a1b5c3a", compress("aabcccccaaa"));
}

@Test
public void test_a5() {
    assertEquals("5a", compress("aaaaa"));
}

@Test
public void test_empty() {
    assertEquals("", compress(""));
}

@Test
public void test_a() {
    assertEquals("1a", compress("a"));
}

@Test
public void test_a3b4() {
    assertEquals("3a4b", compress("aaabbbb"));
}

@Test
public void test_abc() {
    assertEquals("1a1b1c", compress("abc"));
}

@Test
public void test_wwwggopp() {
    assertEquals("3w2g1o2p", compress("wwwggopp"));
}
\$\endgroup\$
1
  • \$\begingroup\$ +1 for your answer . as testCase test_aabcccccaaa fails for the original code , But this scenario is covered by you ... As you suggested Map should not have been used here. I would say that beacause of Map this bug has happened. \$\endgroup\$ Oct 12, 2014 at 10:12
2
\$\begingroup\$

Your program is an answer to a different question. It counts characters, as opposed to characters in runs.

Here is an example of the difference:

Input: "difference".

Output requested by the question: "1d 1i 2f 1e 1r 1e 1n 1c 1e"

Your output: "1d 1i 2f 3e 1r 1n 1c"

(I put the spaces in to make it more readable)

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.