vanitasvitae
/
Smack
mirror of https://github.com/vanitasvitae/Smack.git
1
0
Fork 0
Code Issues Releases Wiki Activity

Support XML_ROUNDTRIP where possible 

The approach to avoid XmlPullParser's XML_ROUNDTRIP feature prevents
Smack from handling XML namespaces on e.g. PubSub payloads on platforms
where support for XML_ROUNDTRIP would be available.

We now check for this XmlPullParser feature and use it if it's
available.
This commit is contained in:
Florian Schmaus 2014-10-15 21:02:16 +02:00
parent 9a6d042652
commit ed58abbee7
1 changed files with 43 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

49
smack-core/src/main/java/org/jivesoftware/smack/util/PacketParserUtils.java
View File

@ -59,6 +59,8 @@ import org.xmlpull.v1.XmlPullParserFactory;
public class PacketParserUtils {
private static final Logger LOGGER = Logger.getLogger(PacketParserUtils.class.getName());
public static final String FEATURE_XML_ROUNDTRIP = "http://xmlpull.org/v1/doc/features.html#xml-roundtrip";
public static XmlPullParser getParserFor(String stanza) throws XmlPullParserException, IOException {
return getParserFor(new StringReader(stanza));
}
@ -144,6 +146,12 @@ public class PacketParserUtils {
public static XmlPullParser newXmppParser() throws XmlPullParserException {
XmlPullParser parser = XmlPullParserFactory.newInstance().newPullParser();
parser.setFeature(XmlPullParser.FEATURE_PROCESS_NAMESPACES, true);
try {
parser.setFeature(FEATURE_XML_ROUNDTRIP, true);
} catch (XmlPullParserException e) {
// Doesn't matter if FEATURE_XML_ROUNDTRIP isn't available
LOGGER.log(Level.FINEST, "XmlPullParser does not support XML_ROUNDTRIP", e);
}
return parser;
}
@ -368,12 +376,14 @@ public class PacketParserUtils {
* not nested ones, if <code>fullNamespaces</code> is false. If it is true, then namespaces of
* parent elements will be added to child elements that don't define a different namespace.
* <p>
* This method is able to parse the content with MX- and KXmlParser. In order to achieve
* this some trade-off has to be make, because KXmlParser does not support xml-roundtrip (ie.
* return a String on getText() on START_TAG and END_TAG). We are therefore required to work
* around this limitation, which results in only partial support for XML namespaces ("xmlns"):
* Only the outermost namespace of elements will be included in the resulting String, if
* <code>fullNamespaces</code> is set to false.
* This method is able to parse the content with MX- and KXmlParser. KXmlParser does not support
* xml-roundtrip. i.e. return a String on getText() on START_TAG and END_TAG. We check for the
* XML_ROUNDTRIP feature. If it's not found we are required to work around this limitation, which
* results in only partial support for XML namespaces ("xmlns"): Only the outermost namespace of
* elements will be included in the resulting String, if <code>fullNamespaces</code> is set to false.
* </p>
* <p>
* In particular Android's XmlPullParser does not support XML_ROUNDTRIP.
* </p>
*
* @param parser
@ -384,6 +394,15 @@ public class PacketParserUtils {
* @throws IOException
*/
public static CharSequence parseContentDepth(XmlPullParser parser, int depth, boolean fullNamespaces) throws XmlPullParserException, IOException {
if (parser.getFeature(FEATURE_XML_ROUNDTRIP)) {
return parseContentDepthWithRoundtrip(parser, depth, fullNamespaces);
} else {
return parseContentDepthWithoutRoundtrip(parser, depth, fullNamespaces);
}
}
private static CharSequence parseContentDepthWithoutRoundtrip(XmlPullParser parser, int depth,
boolean fullNamespaces) throws XmlPullParserException, IOException {
XmlStringBuilder xml = new XmlStringBuilder();
int event = parser.getEventType();
boolean isEmptyElement = false;
@ -437,6 +456,24 @@ public class PacketParserUtils {
return xml;
}
private static CharSequence parseContentDepthWithRoundtrip(XmlPullParser parser, int depth, boolean fullNamespaces)
throws XmlPullParserException, IOException {
StringBuilder sb = new StringBuilder();
int event = parser.getEventType();
outerloop: while (true) {
// Only append the text if the parser is not on on an empty element' start tag. Empty elements are reported
// twice, so in order to prevent duplication we only add their text when we are on their end tag.
if (!(event == XmlPullParser.START_TAG && parser.isEmptyElementTag())) {
sb.append(parser.getText());
}
if (event == XmlPullParser.END_TAG && parser.getDepth() <= depth) {
break outerloop;
}
event = parser.next();
}
return sb;
}
/**
* Parses a presence packet.
*