From 562f224c2136391f1e9ecb044aaa84cdca8a04e7 Mon Sep 17 00:00:00 2001 From: 4pr0n Date: Sun, 20 Jul 2014 01:13:02 -0700 Subject: [PATCH] Added cheeby ripper #8 --- .../ripme/ripper/rippers/CheebyRipper.java | 68 +++++++++++++++++++ 1 file changed, 68 insertions(+) create mode 100644 src/main/java/com/rarchives/ripme/ripper/rippers/CheebyRipper.java diff --git a/src/main/java/com/rarchives/ripme/ripper/rippers/CheebyRipper.java b/src/main/java/com/rarchives/ripme/ripper/rippers/CheebyRipper.java new file mode 100644 index 00000000..c09d615f --- /dev/null +++ b/src/main/java/com/rarchives/ripme/ripper/rippers/CheebyRipper.java @@ -0,0 +1,68 @@ +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 CheebyRipper extends AbstractHTMLRipper { + + public CheebyRipper(URL url) throws IOException { + super(url); + } + + @Override + public String getHost() { + return "cheeby"; + } + @Override + public String getDomain() { + return "cheeby.com"; + } + + @Override + public String getGID(URL url) throws MalformedURLException { + Pattern p = Pattern.compile("^https?://[w.]*cheeby.com/u/([a-zA-Z0-9\\-_]{3,}).*$"); + Matcher m = p.matcher(url.toExternalForm()); + if (m.matches()) { + return m.group(1); + } + throw new MalformedURLException("cheeby user not found in " + url + ", expected http://cheeby.com/u/username"); + } + + @Override + public URL sanitizeURL(URL url) throws MalformedURLException { + return new URL("http://cheeby.com/u/" + getGID(url) + "/pics"); + } + + @Override + public Document getFirstPage() throws IOException { + return Http.url(this.url) + .get(); + } + + @Override + public List getURLsFromPage(Document page) { + List imageURLs = new ArrayList(); + for (Element image : page.select("div.i a img")) { + String imageURL = image.attr("src"); + imageURL = imageURL.replace("s.", "."); + imageURLs.add(imageURL); + } + return imageURLs; + } + + @Override + public void downloadURL(URL url, int index) { + addURLToDownload(url, getPrefix(index)); + } +}