mirror/ripme
1
0
Fork 0
mirror of https://github.com/RipMeApp/ripme.git synced 2025-08-27 15:49:50 +02:00
Code Issues Releases Wiki Activity

Merge pull request #1136 from 0x1f595/jabarchives

Browse Source 
Updating Jab Archives ripper to add image title
This commit is contained in:
cyian-1756
2019-01-04 07:44:47 -05:00
committed by GitHub
parent 46dd89de1b b5fc3a166a
commit 3f70a67860
1 changed files with 25 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

27
src/main/java/com/rarchives/ripme/ripper/rippers/JabArchivesRipper.java
View File

@@ -7,6 +7,12 @@ import java.util.ArrayList;
import java.util.List; import java.util.List;
import java.util.regex.Matcher; import java.util.regex.Matcher;
import java.util.regex.Pattern; import java.util.regex.Pattern;
import java.util.Collections;
import java.util.HashMap;
import java.util.Map;
import java.util.Locale;
import java.text.Normalizer;
import java.text.Normalizer.Form;
import com.rarchives.ripme.ripper.AbstractHTMLRipper; import com.rarchives.ripme.ripper.AbstractHTMLRipper;
import com.rarchives.ripme.utils.Http; import com.rarchives.ripme.utils.Http;
@@ -17,6 +23,11 @@ import org.jsoup.select.Elements;
public class JabArchivesRipper extends AbstractHTMLRipper { public class JabArchivesRipper extends AbstractHTMLRipper {
private static final Pattern NONLATIN = Pattern.compile("[^\\w-]");
private static final Pattern WHITESPACE = Pattern.compile("[\\s]");
private Map<String, String> itemPrefixes = Collections.synchronizedMap(new HashMap<String, String>());
public JabArchivesRipper(URL url) throws IOException { public JabArchivesRipper(URL url) throws IOException {
super(url); super(url);
} }
@@ -62,17 +73,29 @@ public class JabArchivesRipper extends AbstractHTMLRipper {
return Http.url(nextUrl).get(); return Http.url(nextUrl).get();
} }
protected String getSlug(String input) {
// Get a URL/file-safe version of a string
String nowhitespace = WHITESPACE.matcher(input).replaceAll("-");
String normalized = Normalizer.normalize(nowhitespace, Form.NFD);
String slug = NONLATIN.matcher(normalized).replaceAll("");
return slug.toLowerCase(Locale.ENGLISH);
}
@Override @Override
public List<String> getURLsFromPage(Document doc) { public List<String> getURLsFromPage(Document doc) {
List<String> result = new ArrayList<String>(); List<String> result = new ArrayList<String>();
for (Element el : doc.select("#contentMain img")) { for (Element el : doc.select("#contentMain img")) {
result.add("https://jabarchives.com" + el.attr("src").replace("thumb", "large")); String url = "https://jabarchives.com" + el.attr("src").replace("thumb", "large");
result.add(url);
String title = el.parent().attr("title");
itemPrefixes.put(url, getSlug(title) + "_");
} }
return result; return result;
} }
@Override @Override
public void downloadURL(URL url, int index) { public void downloadURL(URL url, int index) {
addURLToDownload(url, getPrefix(index)); addURLToDownload(url, itemPrefixes.get(url.toString()));
} }
} }