Improved multipage image extraction

dafi · Feb 14, 2016 · c84282b · c84282b
1 parent ac63b09
commit c84282b
Show file tree

Hide file tree

Showing 2 changed files with 4 additions and 2 deletions.
diff --git a/app/src/main/assets/domSelectors.json b/app/src/main/assets/domSelectors.json
@@ -86,7 +86,7 @@
     "hotflick.net": {
       "image": ".pict_content img",
       "container": ".box-item-155-img img",
-      "multiPage": ".box-paging a:not(.active)"
+      "multiPage": ".box-paging a.active ~ a"
     },
     "carreck.com": {
       "title": ".postentry p"

diff --git a/app/src/main/java/com/ternaryop/photoshelf/fragment/ImagePickerFragment.java b/app/src/main/java/com/ternaryop/photoshelf/fragment/ImagePickerFragment.java
@@ -298,12 +298,14 @@ private void extractImageFromMultiPage(List<ImageInfo> imageInfoList, DOMSelecto
             if (selector.getMultiPage() == null) {
                 return;
             }
-            for (Element element : startPageDocument.select(selector.getMultiPage())) {
+            Element element = startPageDocument.select(selector.getMultiPage()).first();
+            while (element != null) {
                 String pageUrl = element.absUrl("href");
                 String pageContent = readURLContent(pageUrl);
                 Document pageDocument = Jsoup.parse(pageContent);
                 pageDocument.setBaseUri(pageUrl);
                 extractImages(imageInfoList, domSelectorFinder.getSelectorFromUrl(pageUrl), pageDocument);
+                element = pageDocument.select(selector.getMultiPage()).first();
             }
         }