Add AhoCorasick

src/main/java/com/thealgorithms/strings/AhoCorasick.java

@@ -0,0 +1,227 @@
+/*
+ * Aho-Corasick String Matching Algorithm Implementation
+ *
+ * This code implements the Aho-Corasick algorithm, which is used for efficient
+ * string matching in a given text. It can find multiple patterns simultaneously
+ * and records their positions in the text.
+ *
+ * Author: Prabhat-Kumar-42
+ * GitHub: https://github.com/Prabhat-Kumar-42
+ */
+
+package com.thealgorithms.strings;
+
+import java.util.ArrayList;
+import java.util.HashMap;
+import java.util.LinkedList;
+import java.util.Map;
+import java.util.Queue;
+
+public class AhoCorasick {
+
+    // Trie Node Class
+    public class Node {
+        // Represents a character in the trie
+        private HashMap<Character, Node> child = new HashMap<>(); // Child nodes of the current node
+        private Node suffixLink; // Suffix link to another node in the trie
+        private Node outputLink; // Output link to another node in the trie
+        private int patternInd; // Index of the pattern that ends at this node
+
+        Node() {
+            this.suffixLink = null;
+            this.outputLink = null;
+            this.patternInd = -1;
+        }
+
+        public HashMap<Character, Node> getChild() {
+            return child;
+        }
+
+        public Node getSuffixLink() {
+            return suffixLink;
+        }
+
+        public void setSuffixLink(final Node suffixLink) {
+            this.suffixLink = suffixLink;
+        }
+
+        public Node getOutputLink() {
+            return outputLink;
+        }
+
+        public void setOutputLink(final Node outputLink) {
+            this.outputLink = outputLink;
+        }
+
+        public int getPatternInd() {
+            return patternInd;
+        }
+
+        public void setPatternInd(final int patternInd) {
+            this.patternInd = patternInd;
+        }
+    }
+
+    // Trie Class
+    public class Trie {
+
+        private Node root = null; // Root node of the trie
+        private final String[] patterns; // patterns according to which Trie is constructed
+
+        public Trie(final String[] patterns) {
+            root = new Node(); // Initialize the root of the trie
+            this.patterns = patterns;
+            buildTrie();
+            buildSuffixAndOutputLinks();
+        }
+
+        // builds AcoCorasick Trie
+        private void buildTrie() {
+
+            // Loop through each input pattern and building Trie
+            for (int i = 0; i < patterns.length; i++) {
+                Node curr = root; // Start at the root of the trie for each pattern
+
+                // Loop through each character in the current pattern
+                for (int j = 0; j < patterns[i].length(); j++) {
+                    char c = patterns[i].charAt(j); // Get the current character
+
+                    // Check if the current node has a child for the current character
+                    if (curr.getChild().containsKey(c)) {
+                        curr = curr.getChild().get(c); // Update the current node to the child node
+                    } else {
+                        // If no child node exists, create a new one and add it to the current node's children
+                        Node nn = new Node();
+                        curr.getChild().put(c, nn);
+                        curr = nn; // Update the current node to the new child node
+                    }
+                }
+                curr.setPatternInd(i); // Store the index of the pattern in the current leaf node
+            }
+        }
+
+        private void initializeSuffixLinksForChildNodesOfTheRoot(Queue<Node> q) {
+            for (char rc : root.getChild().keySet()) {
+                Node childNode = root.getChild().get(rc);
+                q.add(childNode); // Add child node to the queue
+                childNode.setSuffixLink(root); // Set suffix link to the root
+            }
+        }
+
+        private void buildSuffixAndOutputLinks() {
+            root.setSuffixLink(root); // Initialize the suffix link of the root to itself
+            Queue<Node> q = new LinkedList<>(); // Initialize a queue for BFS traversal
+
+            initializeSuffixLinksForChildNodesOfTheRoot(q);
+
+            while (!q.isEmpty()) {
+                Node currentState = q.poll(); // Get the current node for processing
+
+                // Iterate through child nodes of the current node
+                for (char cc : currentState.getChild().keySet()) {
+                    Node currentChild = currentState.getChild().get(cc); // Get the child node
+                    Node parentSuffix = currentState.getSuffixLink(); // Get the parent's suffix link
+
+                    // Calculate the suffix link for the child based on the parent's suffix link
+                    while (!parentSuffix.getChild().containsKey(cc) && parentSuffix != root) {
+                        parentSuffix = parentSuffix.getSuffixLink();
+                    }
+
+                    // Set the calculated suffix link or default to root
+                    if (parentSuffix.getChild().containsKey(cc)) {
+                        currentChild.setSuffixLink(parentSuffix.getChild().get(cc));
+                    } else {
+                        currentChild.setSuffixLink(root);
+                    }
+
+                    q.add(currentChild); // Add the child node to the queue for further processing
+                }
+
+                // Establish output links for nodes to efficiently identify patterns within patterns
+                if (currentState.getSuffixLink().getPatternInd() >= 0) {
+                    currentState.setOutputLink(currentState.getSuffixLink());
+                } else {
+                    currentState.setOutputLink(currentState.getSuffixLink().getOutputLink());
+                }
+            }
+        }
+
+        // Searches for patterns in the input text and records their positions
+        public ArrayList<ArrayList<Integer>> searchIn(String text) {
+            /*
+             * positionByStringIndexValue is the list of list containing the indexes of words in patterns/dictionary
+             * list index represents word.
+             * eg - list[0] contains the list of start-index of word pattern[0]
+             *      list[1] contains the list of start-index of word pattern[1]
+             *      ......
+             *      ......
+             *      list[n] contains the list of start-index of word pattern[n]
+             */
+            ArrayList<ArrayList<Integer>> positionByStringIndexValue = new ArrayList<>(patterns.length); // Stores positions where patterns are found in the text
+            Node parent = root; // Start searching from the root node
+
+            for (int i = 0; i < patterns.length; i++) {
+                positionByStringIndexValue.add(new ArrayList<Integer>()); // Initialize a list to store positions of the current pattern
+            }
+
+            for (int i = 0; i < text.length(); i++) {
+                char ch = text.charAt(i); // Get the current character in the text
+
+                // Check if the current node has a child for the current character
+                if (parent.getChild().containsKey(ch)) {
+                    parent = parent.getChild().get(ch); // Update the current node to the child node
+
+                    // If the current node represents a pattern, record its position in positionByStringIndexValue
+                    if (parent.getPatternInd() > -1) {
+                        positionByStringIndexValue.get(parent.getPatternInd()).add(i);
+                    }
+
+                    Node outputLink = parent.getOutputLink();
+                    // Follow output links to find and record positions of other patterns
+                    while (outputLink != null) {
+                        positionByStringIndexValue.get(outputLink.getPatternInd()).add(i);
+                        outputLink = outputLink.getOutputLink();
+                    }
+                } else {
+                    // If no child node exists for the character, backtrack using suffix links
+                    while (parent != root && !parent.getChild().containsKey(ch)) {
+                        parent = parent.getSuffixLink();
+                    }
+                    if (parent.getChild().containsKey(ch)) {
+                        i--; // Decrement i to reprocess the same character
+                    }
+                }
+            }
+            setUpStartPoints(positionByStringIndexValue);
+            return positionByStringIndexValue;
+        }
+        // by default positionByStringIndexValue contains end-points. This function converts those
+        // endpoints to start points
+        private void setUpStartPoints(ArrayList<ArrayList<Integer>> positionByStringIndexValue) {
+            for (int i = 0; i < patterns.length; i++) {
+                for (int j = 0; j < positionByStringIndexValue.get(i).size(); j++) {
+                    int endpoint = positionByStringIndexValue.get(i).get(j);
+                    positionByStringIndexValue.get(i).set(j, endpoint - patterns[i].length() + 1);
+                }
+            }
+        }
+    }
+
+    // method to search for patterns in text
+    public static Map<String, ArrayList<Integer>> search(String text, String[] patterns) {
+        Trie trie = new AhoCorasick().new Trie(patterns);
+        ArrayList<ArrayList<Integer>> positionByStringIndexValue = trie.searchIn(text);
+        return convert(positionByStringIndexValue, patterns);
+    }
+
+    // method for converting results to a map
+    private static Map<String, ArrayList<Integer>> convert(ArrayList<ArrayList<Integer>> positionByStringIndexValue, String[] patterns) {
+        Map<String, ArrayList<Integer>> positionByString = new HashMap<>();
+        for (int i = 0; i < patterns.length; i++) {
+            String pattern = patterns[i];
+            ArrayList<Integer> positions = positionByStringIndexValue.get(i);
+            positionByString.put(pattern, new ArrayList<>(positions));
+        }
+        return positionByString;
+    }
+}

src/test/java/com/thealgorithms/strings/AhoCorasickTest.java

@@ -0,0 +1,119 @@
+/*
+ * Tests For Aho-Corasick String Matching Algorithm
+ *
+ *            Author: Prabhat-Kumar-42
+ *    GitHub: https://github.com/Prabhat-Kumar-42
+ */
+
+package com.thealgorithms.strings;
+
+import static org.junit.jupiter.api.Assertions.*;
+
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Map;
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.Test;
+
+/**
+ * This class contains test cases for the Aho-Corasick String Matching Algorithm.
+ * The Aho-Corasick algorithm is used to efficiently find all occurrences of multiple
+ * patterns in a given text.
+ */
+class AhoCorasickTest {
+    private String[] patterns; // The array of patterns to search for
+    private String text; // The input text to search within
+
+    /**
+     * This method sets up the test environment before each test case.
+     * It initializes the patterns and text to be used for testing.
+     */
+    @BeforeEach
+    void setUp() {
+        patterns = new String[] {"ACC", "ATC", "CAT", "GCG", "C", "T"};
+        text = "GCATCG";
+    }
+
+    /**
+     * Test searching for multiple patterns in the input text.
+     * The expected results are defined for each pattern.
+     */
+    @Test
+    void testSearch() {
+        // Define the expected results for each pattern
+        final var expected = Map.of("ACC", new ArrayList<>(Arrays.asList()), "ATC", new ArrayList<>(Arrays.asList(2)), "CAT", new ArrayList<>(Arrays.asList(1)), "GCG", new ArrayList<>(Arrays.asList()), "C", new ArrayList<>(Arrays.asList(1, 4)), "T", new ArrayList<>(Arrays.asList(3)));
+        assertEquals(expected, AhoCorasick.search(text, patterns));
+    }
+
+    /**
+     * Test searching with an empty pattern array.
+     * The result should be an empty map.
+     */
+    @Test
+    void testEmptyPatterns() {
+        // Define an empty pattern array
+        final var emptyPatterns = new String[] {};
+        assertTrue(AhoCorasick.search(text, emptyPatterns).isEmpty());
+    }
+
+    /**
+     * Test searching for patterns that are not present in the input text.
+     * The result should be an empty list for each pattern.
+     */
+    @Test
+    void testPatternNotFound() {
+        // Define patterns that are not present in the text
+        final var searchPatterns = new String[] {"XYZ", "123"};
+        final var expected = Map.of("XYZ", new ArrayList<Integer>(), "123", new ArrayList<Integer>());
+        assertEquals(expected, AhoCorasick.search(text, searchPatterns));
+    }
+
+    /**
+     * Test searching for patterns that start at the beginning of the input text.
+     * The expected position for each pattern is 0.
+     */
+    @Test
+    void testPatternAtBeginning() {
+        // Define patterns that start at the beginning of the text
+        final var searchPatterns = new String[] {"GC", "GCA", "GCAT"};
+        final var expected = Map.of("GC", new ArrayList<Integer>(Arrays.asList(0)), "GCA", new ArrayList<Integer>(Arrays.asList(0)), "GCAT", new ArrayList<Integer>(Arrays.asList(0)));
+        assertEquals(expected, AhoCorasick.search(text, searchPatterns));
+    }
+
+    /**
+     * Test searching for patterns that end at the end of the input text.
+     * The expected positions are 4, 3, and 2 for the patterns.
+     */
+    @Test
+    void testPatternAtEnd() {
+        // Define patterns that end at the end of the text
+        final var searchPatterns = new String[] {"CG", "TCG", "ATCG"};
+        final var expected = Map.of("CG", new ArrayList<Integer>(Arrays.asList(4)), "TCG", new ArrayList<Integer>(Arrays.asList(3)), "ATCG", new ArrayList<Integer>(Arrays.asList(2)));
+        assertEquals(expected, AhoCorasick.search(text, searchPatterns));
+    }
+
+    /**
+     * Test searching for patterns with multiple occurrences in the input text.
+     * The expected sizes are 1 and 1, and the expected positions are 2 and 3
+     * for the patterns "AT" and "T" respectively.
+     */
+    @Test
+    void testMultipleOccurrencesOfPattern() {
+        // Define patterns with multiple occurrences in the text
+        final var searchPatterns = new String[] {"AT", "T"};
+        final var expected = Map.of("AT", new ArrayList<Integer>(Arrays.asList(2)), "T", new ArrayList<Integer>(Arrays.asList(3)));
+        assertEquals(expected, AhoCorasick.search(text, searchPatterns));
+    }
+
+    /**
+     * Test searching for patterns in a case-insensitive manner.
+     * The search should consider patterns regardless of their case.
+     */
+    @Test
+    void testCaseInsensitiveSearch() {
+        // Define patterns with different cases
+        final var searchPatterns = new String[] {"gca", "aTc", "C"};
+        final var expected = Map.of("gca", new ArrayList<Integer>(), "aTc", new ArrayList<Integer>(), "C", new ArrayList<Integer>(Arrays.asList(1, 4)));
+        assertEquals(expected, AhoCorasick.search(text, searchPatterns));
+    }
+}