Issue #6 Checker. CLI and scoring implemented

import java.util.HashMap;

import java.util.HashSet;

import java.util.List;

import java.util.Map;

import java.util.Set;

public class AnomalyDetector {

    private static final Set<String> commonWords = new HashSet<>();

    private static final int AVERAGE_WORD_LENGTH = 5; // TODO: Change later

    static { // simple list of common words for demonstration purposes

        commonWords.add("the");

        commonWords.add("be");

        commonWords.add("to");

        commonWords.add("of");

        commonWords.add("and"); // we can add more later

    }

    public Map<String, Integer> analyzeSentences(List<String> sentences) {

        Map<String, Integer> sentenceScores = new HashMap<>();

        for (String sentence : sentences) {

            int score = 0;

            // length variance

            score += Math.abs(sentence.length() - AVERAGE_WORD_LENGTH * 10); // Assuming an average sentence length

            // word rarity

            String[] words = sentence.split("\\s+");

            for (String word : words) {

                if (!commonWords.contains(word.toLowerCase())) {

                    score += 10; // Increment score for each UNCOMMON word

                }

            }

            sentenceScores.put(sentence, Math.min(score, 100)); // Limit to 100 (normalize)

        }

        return sentenceScores;

    }

    public Map<String, Integer> analyzePhrases(List<String> sentences) {

        Map<String, Integer> phraseScores = new HashMap<>();

        for (String sentence : sentences) {

            List<String> phrases = TextProcessor.extractPhrases(sentence, 2, 3);

            for (String phrase : phrases) {

                int score = 0;

                // Score length variance like above

                if (phrase.length() < AVERAGE_WORD_LENGTH || phrase.length() > AVERAGE_WORD_LENGTH * 3) {

                    score += 20;

                }

                // Word rarity

                String[] words = phrase.split("\\s+");

                for (String word : words) {

                    if (!commonWords.contains(word.toLowerCase())) {

                        score += 5;

                    }

                }

                phraseScores.put(phrase, Math.min(score, 100)); // Normalize

            }

        }

        return phraseScores;

    }

}

import java.io.IOException;

import java.nio.file.Files;

import java.nio.file.Paths;

public class CLI {

    public static void main(String[] args) {

        // Check if the correct number of arguments are passed

        if (args.length > 1 && "--file".equals(args[0])) {

            String path = args[1];

            try {

                // Read the entire file content into a single String

                String content = new String(Files.readAllBytes(Paths.get(path)));

                // Now, you can analyze this content

                Checker checker = new Checker();

                checker.analyze(content);

            } catch (IOException e) {

                // Handle exceptions, like file not found

                System.err.println("Error reading file: " + e.getMessage());

            }

        } else {

            System.out.println("Invalid arguments. Usage: CLI --file [path]");

        }

    }

}

import java.io.IOException;

import java.nio.file.Files;

import java.nio.file.Paths;

import java.util.List;

import java.util.Map;

public class Checker {

    public void analyze(String text) {

        List<String> sentences = TextProcessor.extractSentences(text);

        AnomalyDetector detector = new AnomalyDetector();

        Map<String, Integer> sentenceScores = detector.analyzeSentences(sentences);

        Map<String, Integer> phraseScores = detector.analyzePhrases(sentences);

        // Output results

        System.out.println("{");

        System.out.println("\"sentences\": " + sentenceScores + ",");

        System.ou

 No newline at end of file

import java.util.ArrayList;

import java.util.HashSet;

import java.util.List;

import java.util.Set;

public class TextProcessor {

    public static List<String> extractSentences(String text) {

        List<String> sentences = List.of(text.split("\\."));

        return new ArrayList<>(sentences);

    }

    // Extracts phrases with variable lengths using n-gram method

    public static List<String> extractPhrases(String sentence, int minN, int maxN) {

        // Using a Set to avoid duplicate phrases

        Set<String> phraseSet = new HashSet<>();

        String[] words = sentence.split("\\s+");

        // Loop over the range of n values

        for (int n = minN; n <= maxN; n++) {

            for (int i = 0; i < words.length - n + 1; i++) {

                StringBuilder sb = new StringBuilder();

                for (int j = i; j < i + n; j++) {

                    sb.append((j > i ? " " : "") + words[j]);

                }

                phraseSet.add(sb.toString());

            }

        }

        // Convert the set back to a list to maintain the original interface

        return new ArrayList<>(phraseSet);

    }

}

This is strange so choice word.

 No newline at end of file