Jawn78

Untitled

Dec 14th, 2018
168
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
Java 3.45 KB | None | 0 0
  1. public class info {
  2.    
  3.     private static final String SAMPLE_CSV_FILE = "./sample.csv";
  4.     public static void main(String[] args) throws InterruptedException, IOException {
  5.  // this was testing from a page saved sourcefile
  6. //File input = new File("C:\\Users\\jonat\\Desktop\\haily0.html");
  7. //Document doc = Jsoup.parse(input , "UTF-8", "C:\\Users\\jonat\\Desktop\\info4.html");
  8.   Document doc = Jsoup.parse("Insert current pagesource here");
  9.          
  10. //System.out.println(bradon.text());
  11.         Element[] person_block;
  12.         Element[] email_block;
  13.         Element[] phone_block;
  14.         Element[]  company_phone_block;
  15.         Element[]  company_name_block;
  16.         Element[] company_detail_block;
  17.         Element[] company_content_left;
  18.         person_block = doc.getElementsByClass("person-row-container").toArray(new Element[doc.getElementsByClass("person-row-container").size()]);
  19. System.out.println(doc.getElementsByClass("contact-details-content").size());        
  20. email_block = doc.getElementsByClass("email-link").toArray(new Element[doc.getElementsByClass("email-link").size()]);
  21. phone_block = doc.getElementsByClass("direct-phone").toArray(new Element[doc.getElementsByClass("direct-phone").size()]);
  22. company_phone_block = doc.getElementsByClass("hq-phone").toArray(new Element[doc.getElementsByClass("hq-phone").size()]);
  23. company_name_block = doc.getElementsByClass("company-name long-text").toArray(new Element[doc.getElementsByClass("company-name long-text").size()]);
  24. company_detail_block = doc.getElementsByClass("primaryCompany-container").toArray(new Element[doc.getElementsByClass("primaryCompany-container").size()]);
  25. company_content_left = doc.getElementsByClass("primary-content ng-star-inserted").toArray(new Element[doc.getElementsByClass("primary-content ng-star-inserted").size()]);
  26. //primary-content-left
  27.  
  28.         int lengthy = person_block.length;
  29.  
  30.  try (
  31.             BufferedWriter writer = Files.newBufferedWriter(Paths.get(SAMPLE_CSV_FILE));
  32.  
  33.             CSVPrinter csvPrinter = new CSVPrinter(writer, CSVFormat.DEFAULT
  34.                     .withHeader("ID", "Full Name", "Title","Company", "Email", "Phone", "Company Phone", "Website","Address"));
  35.         ) {
  36.             int i = 0;
  37.        
  38.      
  39.      while(i < person_block.length){
  40.                 String fullName = person_block[i].getElementsByClass("primary-name").text();
  41.                 String title = person_block[i].getElementsByClass("person-title").text();
  42.                 String email = email_block[i].getElementsByClass("email-link").text().replace("Email", "");
  43.                 String phone = phone_block[i].getElementsByClass("direct-phone").text().replace("- (Direct)","").replace("Phone (Direct)", "").replace("(Direct)", "");
  44.                 String hqPhone = company_phone_block[i].getElementsByClass("hq-phone").text().replace("- (HQ)","").replace("(HQ)","");
  45.                 String company_name = company_name_block[i].getElementsByClass("company-name long-text").text();
  46.                 String company_detail = company_detail_block[i].getElementsByClass("link").text();
  47.                 String location_detail = company_content_left[i].getElementsByClass("local-content location-content").text();
  48.                
  49.                 //fix grabbing html
  50.              csvPrinter.printRecord(i, fullName, title, company_name, email, phone, hqPhone, company_detail, location_detail);
  51.              
  52.             i++;
  53.  }
  54.             csvPrinter.flush();    
  55.    
  56.         }
  57.     }
Advertisement
Add Comment
Please, Sign In to add comment