Archives/Athou_commafeed
1
0
Fork 1
mirror of https://github.com/Athou/commafeed.git synced 2026-03-21 21:37:29 +00:00
Code Issues Packages Projects Releases Wiki Activity

detect encoding and trim invalid characters before parsing xml (#60)

Browse Source
This commit is contained in:
Athou
2013-04-23 07:20:21 +02:00
parent 9cb56557be
commit 64747881de
3 changed files with 42 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
src/main/java/com/commafeed/backend/feeds/FeedParser.java
View File

@@ -1,6 +1,7 @@
package com.commafeed.backend.feeds;
import java.io.ByteArrayInputStream;
import java.io.StringReader;
import java.util.Calendar;
import java.util.Date;
import java.util.List;
@@ -41,14 +42,12 @@ public class FeedParser {
feed.setLastUpdated(Calendar.getInstance().getTime());
try {
InputSource source = new InputSource(new ByteArrayInputStream(xml));
if (new String(ArrayUtils.subarray(xml, 0, 100))
.split(SystemUtils.LINE_SEPARATOR)[0].toUpperCase()
.contains("ISO-8859-1")) {
// they probably use word, we need to handle curly quotes and
// other word special characters
source.setEncoding("windows-1252");
}
String encoding = FeedUtils.guessEncoding(xml);
String xmlString = FeedUtils.trimInvalidXmlCharacters(new String(
xml, encoding));
InputSource source = new InputSource(new StringReader(xmlString));
SyndFeed rss = new SyndFeedInput().build(source);
feed.setUrl(feedUrl);
feed.setTitle(rss.getTitle());