Added basic pichunter ripper

This commit is contained in:
cyian-1756 2017-11-16 13:58:21 -05:00
parent c94de79c46
commit 149ab50dc1

View File

@ -0,0 +1,74 @@
package com.rarchives.ripme.ripper.rippers;
import java.io.IOException;
import java.net.MalformedURLException;
import java.net.URL;
import java.util.ArrayList;
import java.util.List;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import com.rarchives.ripme.ripper.AbstractHTMLRipper;
import com.rarchives.ripme.utils.Http;
public class PichunterRipper extends AbstractHTMLRipper {
public PichunterRipper(URL url) throws IOException {
super(url);
}
@Override
public String getHost() {
return "pichunter";
}
@Override
public String getDomain() {
return "pichunter.com";
}
@Override
public String getGID(URL url) throws MalformedURLException {
Pattern p = Pattern.compile("https?://www.pichunter.com/(tags|models|sites)/([a-zA-Z0-9_-]+)/?");
Matcher m = p.matcher(url.toExternalForm());
if (m.matches()) {
return m.group(2);
}
throw new MalformedURLException("Expected pichunter URL format: " +
"pichunter.com/(tags|models|sites)/Name/ - got " + url + " instead");
}
@Override
public Document getFirstPage() throws IOException {
// "url" is an instance field of the superclass
return Http.url(url).get();
}
@Override
public Document getNextPage(Document doc) throws IOException {
// We use comic-nav-next to the find the next page
Element elem = doc.select("div.paperSpacings > ul > li.arrow").last();
String nextPage = elem.select("a").attr("href");
// Some times this returns a empty string
// This for stops that
return Http.url("http://www.pichunter.com" + nextPage).get();
}
@Override
public List<String> getURLsFromPage(Document doc) {
List<String> result = new ArrayList<>();
for (Element el : doc.select("div.thumbtable > a.thumb > img")) {
result.add(el.attr("src").replaceAll("_i", "_o"));
}
return result;
}
@Override
public void downloadURL(URL url, int index) {
addURLToDownload(url, getPrefix(index));
}
}