Appie-Bonus-Scraper/src/main/java/nl/herpiederpiee/appie_scraper/BonusManager.java

package nl.herpiederpiee.appie_scraper;

import com.microsoft.playwright.*;
import xyz.nextn.levenshteindistance.LevenshteinDistance;

import java.text.Normalizer;
import java.util.ArrayList;
import java.util.concurrent.TimeUnit;

public class BonusManager {
    static ArrayList<BonusItem> bonusItems = new ArrayList<BonusItem>();;

    public static void updateBonusItems(){
        try (Playwright playwright = Playwright.create()) {

            Browser browser = playwright.chromium().launch(new BrowserType.LaunchOptions().setHeadless(false));

            BrowserContext context = browser.newContext();
            Page bonusPagina = context.newPage();


            bonusPagina.navigate("https://www.ah.nl/bonus");
            TimeUnit.SECONDS.sleep(5); // wait for page to actaully fully load

            Locator bonusElements = bonusPagina.locator(".promotion-card_root__tQA3z");
            for (ElementHandle bonusElement : bonusElements.elementHandles()){
                BonusItem bonusItem = new BonusItem(bonusElement);

                // exclude annoying elements
                if (bonusItem.category.equals( "onlineOnly")) continue;
                if (bonusItem.category.equals( "gall")) continue;
                if (bonusItem.category.equals( "gall-card")) continue;
                if (bonusItem.category.equals( "etos")) continue;

                bonusItems.add(bonusItem);
                System.out.println("added bonusItem " + bonusItem.title);
            }

        } catch (InterruptedException e) {
            throw new RuntimeException(e);
        }
    }

    public static ArrayList<BonusItem> getBonusItems(String name){
        ArrayList<Pair<BonusItem, Integer>> list = new ArrayList<>();

        if (name == null || name.trim().isEmpty()){
            return bonusItems;
        }

        for (BonusItem bonusItem : bonusItems) {
            Integer score = fuzzyMatchScore(name, bonusItem.title);
            list.add(Pair.pair(bonusItem, score));
        }
        list.sort((a, b) -> Integer.compare(b.second, a.second));

        if (list.get(0).second.equals(0)){
            return new ArrayList<>();
        }
        ArrayList<BonusItem> top10 = new ArrayList<>();
        int i = 0;
        while (top10.size() < 10) {
            top10.add(list.get(i).first);
            i++;
        }

        return top10;
    }

    public static int fuzzyMatchScore(String query, String title) {
        if (query == null || title == null || query.isEmpty() || title.isEmpty()) {
            return 0;
        }

        String normalizedQuery = normalize(query);
        String normalizedTitle = normalize(title);

        // ===== TIER 1: EXACT WORD MATCH (highest priority) =====
        if (isExactWordMatch(normalizedTitle, normalizedQuery)) {
            return 100;
        }

        // ===== TIER 2: WORD-BOUNDARY SUBSTRING =====
        if (isWordBoundaryMatch(normalizedTitle, normalizedQuery)) {
            return 95;
        }

        // ===== TIER 3: PREFIX MATCH =====
        if (isPrefixMatch(normalizedTitle, normalizedQuery)) {
            return 85;
        }

        // ===== TIER 4: LEVENSHTEIN (typo tolerance) =====
        int qlen = normalizedQuery.length();
        int tlen = normalizedTitle.length();

        if (qlen > tlen) {
            return 0;
        }

        int bestDistance = Integer.MAX_VALUE;

        for (int i = 0; i <= tlen - qlen; i++) {
            String sub = normalizedTitle.substring(i, i + qlen);
            int dist = levenshteinDistance(normalizedQuery, sub);
            if (dist < bestDistance) {
                bestDistance = dist;
                if (dist == 0) break;
            }
        }

        // Allow up to 2 edits (typo tolerance)
        if (bestDistance <= 2) {
            // Distance 0 = 80, Distance 1 = 70, Distance 2 = 60
            int score = 80 - (bestDistance * 10);
            return Math.max(0, score);
        }

        return 0;
    }

    /**
     * Exact word match: query must be surrounded by word boundaries or string edges
     * "LOR" matches "L OR" or "LOR coffee" but NOT "LOREAL"
     */
    private static boolean isExactWordMatch(String title, String query) {
        String[] words = title.split("\\s+");
        for (String word : words) {
            if (word.equals(query)) {
                return true;
            }
        }
        return false;
    }

    /**
     * Word boundary match: query matches at word start/end
     * "LOR" matches in "L'OR" (after special char removed)
     * "REAL" matches in "LOREAL" as word boundary? No, stays in Tier 4
     */
    private static boolean isWordBoundaryMatch(String title, String query) {
        // Check if query appears after space or at start
        if (title.startsWith(query + " ")) {
            return true;
        }

        if (title.contains(" " + query)) {
            return true;
        }

        // Check if query ends at word boundary
        if (title.endsWith(" " + query)) {
            return true;
        }

        return false;
    }

    /**
     * Prefix match: query is the start of any word
     * "CAF" matches in "CAFFE" or "CAFE LATTE"
     */
    private static boolean isPrefixMatch(String title, String query) {
        for (String word : title.split("\\s+")) {
            if (word.startsWith(query) && word.length() > query.length()) {
                return true;
            }
        }
        return false;
    }

    private static String normalize(String input) {
        if (input == null || input.isEmpty()) {
            return input;
        }

        // Remove diacritics
        String decomposed = Normalizer.normalize(input, Normalizer.Form.NFD);
        String withoutDiacritics = decomposed
                .replaceAll("\\p{InCombiningDiacriticalMarks}+", "");

        // Lowercase, remove special chars, normalize spaces
        String cleaned = withoutDiacritics
                .toLowerCase()
                .replaceAll("[^a-z0-9\\s]", "")
                .replaceAll("\\s+", " ")
                .trim();

        return cleaned;
    }

    private static int levenshteinDistance(String query, String title) {
        int qlen = query.length();
        int tlen = title.length();

        if (qlen == 0) return tlen;
        if (tlen == 0) return qlen;

        int[][] dp = new int[qlen + 1][tlen + 1];

        for (int i = 0; i <= qlen; i++) dp[i][0] = i;
        for (int j = 0; j <= tlen; j++) dp[0][j] = j;

        for (int i = 1; i <= qlen; i++) {
            for (int j = 1; j <= tlen; j++) {
                if (query.charAt(i - 1) == title.charAt(j - 1)) {
                    dp[i][j] = dp[i - 1][j - 1];
                } else {
                    dp[i][j] = 1 + Math.min(
                            Math.min(dp[i - 1][j - 1], dp[i - 1][j]),
                            dp[i][j - 1]
                    );
                }
            }
        }

        return dp[qlen][tlen];
    }
}