recept/recept/src/main/java/parser/Koket.java

package parser;

import java.awt.image.BufferedImage;
import java.io.File;
import java.io.IOException;
import java.net.MalformedURLException;
import java.net.URL;
import java.util.List;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

import javax.imageio.ImageIO;

import org.imgscalr.Scalr;
import org.openqa.selenium.By;
import org.openqa.selenium.NoSuchElementException;
import org.openqa.selenium.WebElement;
import org.openqa.selenium.support.ui.ExpectedConditions;

import obejcts.Ingredient;
import obejcts.Recepie;

public class Koket extends ParserBase {
	public static final String RECEPIE_TITLE_XPATH = ".//h2[contains(@class,'list_item_title')]";
	public static final String BASE_URL = "https://www.koket.se/";
	public static final String PATH_SEPARATOR = "%20";
	public static final String RECEPIELIST_ITEMS_XPATH = "//div[contains(@class,'list_item_wrapper')]";
	public static final String URL_SEARCH_PATTERN = "search?searchtext=";
	public static final String COOKIE_CONSENT_BUTTON_XPATH = "//button[@id='onetrust-accept-btn-handler']";

	public static final String IMAGE_XPATH = "//img[contains(@class,'image__')]";

	public static final String SHOW_MORE_XPATH = "//div[contains(@class,'show-more-button')]";

	public static boolean isFromHereRecepie(String url) {
		return url.contains(BASE_URL);
	}

	@Override protected String getAmountFromText(String text) {
		Pattern pattern = Pattern.compile("^[0-9, \\/-]*");
		Matcher matcher = pattern.matcher(text);

		if (matcher.find()) {
			return matcher.group();
		}
		return "";
	}

	@Override void handleRecepie(WebElement recepie, String recepieTitle, Recepie recepieFromDb) {
		if (checkIfElementExists(recepie, ".//div[contains(@class,'rating_wrapper')]")) {
			super.handleRecepie(recepie, recepieTitle, recepieFromDb);
		}
	}

	@Override protected void parseRecepie(WebElement recepie, String recepieTitle) {
		try {
			Thread.sleep(1000);
			recepie.click();
			Thread.sleep(200);

			wait.until(ExpectedConditions.numberOfElementsToBeMoreThan(By.xpath("//h1[contains(@class, 'recipe_title')]"), 0));

			String title = driver.findElement(By.xpath("//h1[contains(@class, 'recipe_title')]")).getText();

			String time = "";
			if (checkIfElementExists(driver, "//div[contains(@class, 'details_wrapper')]//p[2]//span")) {
				String timeText = driver.findElement(By.xpath("//div[contains(@class, 'details_wrapper')]//p[2]//span"))
						.getText();
				time = timeText.replaceAll("[^+0-9]", "");
			}

			Recepie newRecepie = new Recepie();
			newRecepie.setTime(time);
			newRecepie.setUrl(driver.getCurrentUrl());
			newRecepie.setName(recepieTitle);
			WebElement desc = driver.findElement(By.xpath("//div[contains(@class, 'description_description')]"));

			scrollElementIntoView(driver, desc);
			if (checkIfElementExists(desc, ".//p")) {
				newRecepie.setDescription(desc.findElement(By.xpath(".//p")).getText());
			}

			newRecepie.setId(database.insertRecepie(newRecepie));

			List<WebElement> ingredients = driver.findElements(By.xpath("//span[@class='ingredient']//span"));
			String measurement = "";
			String amount = "";
			for (WebElement ingredient : ingredients) {
				String name = fixName(ingredient.getText());
				measurement = getMeasurement(name);
				amount = getAmountFromText(name);
				name = name.replace(measurement, "");
				name = name.replace(amount, "");
				name = name.trim();

				database.addIngredientToRecepie(newRecepie.getId(), new Ingredient(fixName(name), amount, measurement));
			}

			List<WebElement> steps = driver
					.findElements(By.xpath("//ol[contains(@class, 'instruction_section_numberedList')]//span"));

			int stepCounter = 1;
			for (WebElement step : steps) {
				database.addRecepieStep(newRecepie.getId(), stepCounter++, step.getText());
			}

			saveImage(newRecepie);
			driver.navigate().back();

		} catch (InterruptedException e) {
			e.printStackTrace();
		}
	}

	void saveImage(Recepie newRecepie) {

		try {
			WebElement imageElement = driver.findElement(By.xpath(IMAGE_XPATH));
			String imgSrc = imageElement.getAttribute("currentSrc");
			if (imgSrc.contains(".webp")) {
				imgSrc = imgSrc.replace(".webp", "");
			}
			URL imageUrl = new URL(imgSrc);
			BufferedImage savedImage = ImageIO.read(imageUrl);

			BufferedImage resizedImage = Scalr.resize(savedImage, 300);

			File file = new File("recepie-image.png");
			ImageIO.write(resizedImage, "png", file);

			database.saveRecepieImage(newRecepie.getId(), file);

		} catch (MalformedURLException e) {
			e.printStackTrace();
		} catch (IOException e) {
			e.printStackTrace();
		} catch (NoSuchElementException ne) {
			File file = new File("empty.png");
			try {
				file.createNewFile();
			} catch (IOException e) {
				e.printStackTrace();
			}
			database.saveRecepieImage(newRecepie.getId(), file);
		}
	}

}