Is this a logic error Or something else?

Question

Here's what the teacher assigned us:

Suppose we are an online service that provides a bulletin board for its users. We would like to give our users the option of filtering out profanity. We will consider the words cat, dog, and llama to be profane. Write a program that reads a string from the keyboard and tests whether the string contains one of our profane words. Your program should find words like cAt that differ only in case. You must also not identify words that simply contain what might otherwise be considered a profane word. For example, Dogmatic concatenation is a small category should not be considered profane. Allow the user to use the following punctuation: ( , . ? " ' ( ) ! : ;) This will mean that you would be expected to find “The “Cat” is not a doggone llamaman.” or “Cat, and dog can not be llama.” (Note: You will only be responsible for the first occurrence of a given profane word in a sentence. However, more than one profane word may be contained in a sentence. So “Concatenate the cats”, would not find a profane word but “The doggone cat, and dog are not a llama.” would return 2 profane words cat and llama)

so, I tried this code:

import java.util.Scanner;
public class Degrees 
{
    private static Scanner keyboard = new Scanner(System.in);
    public static void main(String[]args)
    {
        System.out.println("Enter a sentence");
        String sentence = keyboard.nextLine();
        sentence = sentence.toLowerCase();
        if(sentence.indexOf("cat ") != -1)
            System.out.println("the profane word cat was detected");
        else
            System.out.println("the profane word cat wasn't detected");

        if(sentence.indexOf("dog ") != -1)
            System.out.println("the profane word dog was detected");
        else
            System.out.println("the profane word dog wasn't detected");
        if(sentence.indexOf("llama ") != -1)
            System.out.println("the profane word llama was detected");
        else
            System.out.println("the profane word llama wasn't detected");


    }       

}

however the code isn't working how it should be. if i wrote "dogmatic dog" it should only check the first occurrence of dog and see that it is within a word and then ignore the second dog. My code is just ughh. I don't know what I'm missing and what I should add. I've been going at this for 6 hours straight I swear. Please help. I just can't think of anything else. So please, I am open to suggestions and hints.

I also tried using a switch statement but for some reason it was only executing the default.

Answer 1

You'll have to create a "mini-parser" that'll iterate the words in the sentence and will check each one of them if it's considered profane or not.

Partially implemented solution:

public static void main(String[] args) {

    String s = "The doggone cat, and dOg are not a llama.";
    s = s.toLowerCase();
    Scanner sc = new Scanner(s);
    List<String> profaneWords = generateProfaneList();
    int counter = 0;
    while (sc.hasNext()) {
        String word = sc.next();
        for (String profane : profaneWords) {
            if (word.matches(".*\\b" + profane + "\\b.*") && // check an exact match
                    ! s.matches(".*" + profane + "[a-z].*\\b" + profane + "\\b.*") && // check that profane is not
                    ! s.matches(".*[a-z]" + profane + ".*\\b" + profane + "\\b.*")) { // included as part of another word
                counter++;
                System.out.println("The word '" + profane + "' is profane!");
            }
        }
    }
    System.out.println(counter + " profane words were found");
}

private static List<String> generateProfaneList() {
    List<String> profaneWords =  new ArrayList<>();
    profaneWords.add("dog");
    profaneWords.add("cat");
    profaneWords.add("llama");
    return profaneWords;
}

OUTPUT

The word 'cat' is profane!
The word 'llama' is profane!
2 profane words were found

Answer 2

I suggest using this algorithm:

• Define all the profane words in an array. Let's call it profaneWords.
• Split the sentence into several strings using whitespace. This will be stored into an array, let's call it wordsToAnalyze
• For each word (string) in profaneWords, let's call the current word profane:
  • Create a flag to check if profane has been found. Let's call it found. Initialize it with a value of no.
  • For each word (string) in wordsToAnalyze, let's call the current word analyzeMe:
    • Trim all non-characters from analyzeMe.
    • Check if analyzeMe is equal to profane. If it does, then mark found to yes and break the for loop.
    • Check if analyzeMe contains profane. If it does, then break the current for loop.
  • If found is yes, then report that the profane word has been identified.

I won't provide the proper Java implementation for algorithm above. Instead, just a pseudo code (after all, it's homework, so it's your job to do the code, not ours =) ):

profaneWords = { "cat", "dog", "llama" } //why llama is profane? =(
wordsToAnalyze = sentence.split(" ") //this can be improved but you should not use regex yet
for each profane in profaneWords
begin for
    found = false
    for each analyzeMe in wordsToAnalyze
    begin for
        analyzeMe = trimNonCharacters(analyzeMe)
        if (analyzeMe is equal to profane)
            found = true
            break
        if (analyzeMe contains profane)
            break
    end for
    if (found is true)
        print "The word " + profane + " was found."
end for

For trimNonCharacters you may create another method that basically reads every character from the string parameter and removes any non-character in it and create a new string. You may use a StringBuilder for this:

public static String trimNonCharacters(String string) {
    int startIndex = 0;
    int endIndex = string.length();
    for (int i = 0; i < string.length(); i++) {
        if (Character.isLetter(string.charAt(i))) {
            break;
        }
        startIndex++;
    }
    for (int i = string.length() - 1; i >= 0; i--) {
        if (Character.isLetter(string.charAt(i))) {
            break;
        }
        endIndex--;
    }
    String result = "";
    if (startIndex <= endIndex) {
        result = string.substring(startIndex, endIndex);
    }
    return result;
}

Answer 3

This is a great candidate for a regex:

System.out.println("Enter a sentence");
String sentence = keyboard.nextLine();
sentence = sentence.toLowerCase();

Pattern p = Pattern.compile("\Wcat\W|\Wdog\W|\Wllama\W");
Matcher m = p.matcher(sentence);
boolean matchFound = m.matches();

\W will match any non-digit and non-word so as an example concatenate would not trigger a match but "cat would.

For more information: http://docs.oracle.com/javase/7/docs/api/java/util/regex/Pattern.html