import org.apache.poi.openxml4j.exceptions.InvalidFormatException;
import org.apache.poi.ss.usermodel.*;
import org.apache.poi.xssf.usermodel.XSSFWorkbook;
import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;

import java.io.File;
import java.io.FileOutputStream;
import java.io.IOException;
import java.net.URL;

public class Main1 {
    public static void main(String[] args) throws IOException, InvalidFormatException {
        // Load the Excel file
        String excelFilePath = "E:\\desktop\\ziku\\bushoubihua.xlsx";
        Workbook workbook = new XSSFWorkbook(new File(excelFilePath));
        Sheet sheet = workbook.getSheetAt(0); // assuming you want the first sheet
        int index = 0;


        for (Row row : sheet) {
            Cell cell = row.getCell(0); // get the first column
            String keyword = cell.getStringCellValue();

            // Replace the keyword in the URL
            String url = "https://zd.hwxnet.com/search.do?keyword=" + keyword;
            Document document = Jsoup.parse(new URL(url),30000);
            Elements elements = document.getElementsByClass("spwid80");
            Elements elements1 = document.getElementsByClass("clearfix");
            Elements elements2 = document.getElementsByClass("spwid120");
            // Add the obtained text to the second column of the current row in the Excel file
            Cell outputCell = row.createCell(1);
            //部首
            if(elements2.get(0).text().equals("通用字")||elements2.get(0).text().equals("通用字、常用字")||elements2.get(0).text().equals("通用字、次常用字")){
                outputCell.setCellValue(elements.get(0).text());
                //总笔画
                Cell outputCell1 = row.createCell(2);
                outputCell1.setCellValue(elements.get(2).text());
                //拼音
                Cell outputCell2 = row.createCell(3);
                String html = elements1.toString();
                Document parse = Jsoup.parse(html);
                Element introduce1 = parse.getElementsByClass("introduce").first();
                // 遍历所有的"introduce"元素
                if (introduce1 != null){
                    // 获取每个"introduce"元素下的"label"元素
                    Elements labels = introduce1.getElementsByClass("label");
                    // 检查是否有足够多的"label"元素
                    if (labels.size() > 1) {
                        // 获取第二个"label"元素
                        Element secondLabel = labels.get(1);
                        // 获取该"label"元素下所有的"span"元素
                        Elements spans = secondLabel.getElementsByTag("span");
                        // 遍历所有的"span"元素并打印其内容
                        String pinyin = "";
                        for (Element span : spans) {
                            pinyin+= span.text()+"\t";
                        }
                        outputCell2.setCellValue(pinyin);
                    }
                }


                // Print a message to the console
                System.out.println("Successfully parsed and wrote keyword: " + keyword);
            }else {
                index++;
//                System.out.println(keyword);
            }
//            System.out.println("抛出"+index+"个");
        }

        // Save the changes to a new file
        String outputExcelFilePath = "E:\\desktop\\ziku\\bushoubihua001.xlsx";
        try (FileOutputStream outputStream = new FileOutputStream(outputExcelFilePath)) {
            workbook.write(outputStream);
        }

        workbook.close();
    }
}