제주도렌트카업체 정보 크롤링 기능 추가

pom.xml

@@ -136,6 +136,13 @@
 			<version>1.0.0</version>
 		</dependency>
 
+
+		<dependency>
+		    <groupId>org.jsoup</groupId>
+		    <artifactId>jsoup</artifactId>
+		    <version>1.21.2</version>
+		</dependency>
+
 		<dependency>
 			<groupId>cokr.xit.base</groupId>
   			<artifactId>xit-crypto</artifactId>

src/main/java/externalsystem/testpool/web/TestPoolController.java

@@ -1,11 +1,21 @@
 package externalsystem.testpool.web;
 
+import java.util.ArrayList;
+import java.util.List;
+
 import javax.annotation.Resource;
 
+import org.jsoup.Jsoup;
+import org.jsoup.nodes.Document;
+import org.jsoup.nodes.Element;
+import org.jsoup.nodes.TextNode;
+import org.jsoup.select.Elements;
 import org.springframework.stereotype.Controller;
 import org.springframework.web.bind.annotation.RequestMapping;
+import org.springframework.web.bind.annotation.ResponseBody;
 import org.springframework.web.servlet.ModelAndView;
 
+import cokr.xit.foundation.data.DataObject;
 import cokr.xit.foundation.web.AbstractController;
 import externalsystem.testpool.RandomUtil;
 import externalsystem.testpool.dao.TestPoolMapper;
@@ -43,4 +53,103 @@ public class TestPoolController extends AbstractController {
 		return mav;
 	}
 
+	@RequestMapping(name="제주도 렌트카 업체 정보 크롤링", value="/jejuRent.do")
+	public ModelAndView jejuRent() throws Exception {
+
+		ModelAndView mav = new ModelAndView("jsonView");
+
+		String site = "https://bizno.net";
+		String listSearch = "/상호명으로사업자등록번호조회";
+		List<String> detailUrls = new ArrayList<String>();
+		List<DataObject> rentEntInfos = new ArrayList<DataObject>();
+
+		Document XM = Jsoup.connect(site+listSearch)
+				.data("area","제주","query","렌트카")
+				.get();
+		Elements titles1 = XM.getElementsByTag("body").get(0).getElementsByClass("titles");
+		for(Element title : titles1) {
+			String detailUrl = site+title.getElementsByTag("a").get(0).attr("href");
+			detailUrls.add(detailUrl);
+		}
+
+		Document XJ = Jsoup.connect(site+listSearch)
+				.data("area","제주","query","렌터카")
+				.get();
+		Elements titles2 = XJ.getElementsByTag("body").get(0).getElementsByClass("titles");
+		for(Element title : titles2) {
+			String detailUrl = site+title.getElementsByTag("a").get(0).attr("href");
+			detailUrls.add(detailUrl);
+		}
+
+		for(String detailUrl : detailUrls) {
+			Document detail = Jsoup.connect(detailUrl).get();
+			String entName = detail.getElementsByTag("h1").get(0).text();
+			Element table = detail.getElementsByClass("table_guide01").get(0);
+
+			String sa = table.getElementsContainingOwnText("사업자등록번호").next().text();
+			String beob = table.getElementsContainingOwnText("법인등록번호").next().text();
+			String tel = table.getElementsContainingOwnText("전화번호").next().text();
+			String fax = table.getElementsContainingOwnText("팩스번호").next().text();
+			String mail = table.getElementsContainingOwnText("회사이메일").next().text();
+			String daepyo = table.getElementsContainingOwnText("대표자명").next().text();
+			String postNo = table.getElementsContainingOwnText("우편번호").next().text();
+			String juso = table.getElementsContainingOwnText("회사주소").next().text();
+
+
+			DataObject rentEntInfo = new DataObject();
+			rentEntInfo.put("기업명", entName);
+			rentEntInfo.put("사업자등록번호", sa);
+			rentEntInfo.put("법인등록번호", beob);
+			rentEntInfo.put("전화번호", tel);
+			rentEntInfo.put("팩스번호", fax);
+			rentEntInfo.put("회사이메일", mail);
+			rentEntInfo.put("대표자명", daepyo);
+			rentEntInfo.put("우편번호", postNo);
+			rentEntInfo.put("회사주소", juso);
+
+			rentEntInfos.add(rentEntInfo);
+		}
+		mav.addObject("result", rentEntInfos);
+
+		return mav;
+	}
+
+	@RequestMapping(name="(주)제주렌트카 정보 크롤링", value="/jejuRent2.do")
+	@ResponseBody
+	public String jejuRent2() throws Exception {
+
+		Document detail = Jsoup.connect("https://bizno.net/article/6168101496").get();
+		String entName = detail.getElementsByTag("h1").get(0).text();
+		Element table = detail.getElementsByClass("table_guide01").get(0);
+
+		String sa = table.getElementsContainingOwnText("사업자등록번호").next().text();
+		String beob = table.getElementsContainingOwnText("법인등록번호").next().text();
+		String tel = table.getElementsContainingOwnText("전화번호").next().text();
+		String fax = table.getElementsContainingOwnText("팩스번호").next().text();
+		String mail = table.getElementsContainingOwnText("회사이메일").next().text();
+		String daepyo = table.getElementsContainingOwnText("대표자명").next().text();
+		String postNo = table.getElementsContainingOwnText("우편번호").next().text();
+
+		List<TextNode> tns = table.getElementsContainingOwnText("회사주소").next().textNodes();
+		List<String> juso = new ArrayList<String>();
+		for(TextNode tn : tns) {
+			juso.add(tn.text());
+		}
+
+
+		DataObject rentEntInfo = new DataObject();
+		rentEntInfo.put("기업명", entName);
+		rentEntInfo.put("사업자등록번호", sa);
+		rentEntInfo.put("법인등록번호", beob);
+		rentEntInfo.put("전화번호", tel);
+		rentEntInfo.put("팩스번호", fax);
+		rentEntInfo.put("회사이메일", mail);
+		rentEntInfo.put("대표자명", daepyo);
+		rentEntInfo.put("우편번호", postNo);
+		rentEntInfo.put("회사주소", juso);
+
+		return toJson(rentEntInfo);
+
+	}
+
 }