From 3bf414c82b1c29733c4a6a0bd751874e6352c901 Mon Sep 17 00:00:00 2001 From: Athou Date: Tue, 26 Mar 2013 22:27:24 +0100 Subject: [PATCH] remove xml tidying as it does not work yet --- pom.xml | 5 ----- .../com/commafeed/backend/feeds/FeedParser.java | 16 +++------------- 2 files changed, 3 insertions(+), 18 deletions(-) diff --git a/pom.xml b/pom.xml index f3fb1104..f6b8fddc 100644 --- a/pom.xml +++ b/pom.xml @@ -172,11 +172,6 @@ jsoup 1.7.2 - - net.sourceforge.htmlcleaner - htmlcleaner - 2.4 - com.google.code.gson diff --git a/src/main/java/com/commafeed/backend/feeds/FeedParser.java b/src/main/java/com/commafeed/backend/feeds/FeedParser.java index d183a26a..1b70680f 100644 --- a/src/main/java/com/commafeed/backend/feeds/FeedParser.java +++ b/src/main/java/com/commafeed/backend/feeds/FeedParser.java @@ -8,10 +8,6 @@ import javax.ejb.Stateless; import org.apache.commons.lang.StringUtils; import org.apache.commons.lang.SystemUtils; -import org.htmlcleaner.CleanerProperties; -import org.htmlcleaner.HtmlCleaner; -import org.htmlcleaner.SimpleHtmlSerializer; -import org.htmlcleaner.TagNode; import org.jsoup.Jsoup; import com.commafeed.backend.model.Feed; @@ -50,6 +46,7 @@ public class FeedParser { feed.getEntries().add(entry); } } catch (Exception e) { + e.printStackTrace(); throw new FeedException(String.format( "Could not parse feed from %s : %s", feedUrl, e.getMessage()), e); @@ -76,15 +73,8 @@ public class FeedParser { } private String balanceTags(String xml) throws Exception { - HtmlCleaner cleaner = new HtmlCleaner(); - CleanerProperties props = cleaner.getProperties(); - props.setOmitXmlDeclaration(true); - TagNode node = cleaner.clean(xml); - SimpleHtmlSerializer serializer = new SimpleHtmlSerializer( - cleaner.getProperties()); - String result = serializer.getAsString(node); - result = StringUtils.trim(xml); - return result; + // TODO close tags + return xml; } private String handleContent(String content) {