I'm trying to deserialize this snippet of XML in Java:
<anime id="16986">
<info type="Picture" src="http://~.jpg" width="141" height="200">
<img src="http://~" width="141" height="200"/>
<img src="http://~" width="318" height="450"/>
</info>
<info type="Main title" lang="EN">Long Riders!</info>
<info type="Alternative title" lang="JA">ろんぐらいだぁす!</info>
</anime>
The problem I'm running into is that the info
element either can have an inline list of img
's or it can just contain text. I was thinking of treating info
as an @Element
in my AnimeHolder class, but I can't have duplicate annotations. I would also like to access the lang
attribute of info to check if it is EN or JP.
I am using these classes to hold the deserialized data:
@Root(name="anime", strict=false)
public class AnimeHolder {
@Attribute(name="id")
private String ANNID;
@ElementList(inline=true)
private List<InfoHolder> infoList;
public String getANNID() {
return ANNID;
}
public List<InfoHolder> getInfoList() {
return infoList;
}
}
and for the info items:
@Root(name="info", strict = false)
public class InfoHolder {
@ElementList(inline=true, required = false)
private List<ImgHolder> imgList;
@Attribute(name = "lang", required = false)
private String language;
public List<ImgHolder> getImgList() {
return imgList;
}
}
With Andreas' I found out I needed to look into handling mixed content. Doing some searching lead me to this solution about creating a custom Converter
. After writing my own and finding out it wasn't being called, this helped sort it out. Here is my re-worked InfoHolder
class and converter:
@Root(name="info", strict = false)
@Convert(InfoHolder.InfoConverter.class)
public class InfoHolder {
private String englishTitle;
private String imageURL;
static class InfoConverter implements Converter<InfoHolder> {
@Override
public InfoHolder read(InputNode node) throws Exception {
String value = node.getValue();
InfoHolder infoHolder = new InfoHolder();
if (value == null){
InputNode nextNode = node.getNext();
while (nextNode != null){
String tag = nextNode.getName();
if (tag.equals("img") && nextNode.getAttribute("src") != null){
infoHolder.imageURL = nextNode.getAttribute("src").getValue();
}
nextNode= node.getNext();
}
} else {
while (node != null){
if (node.getAttribute("lang") != null){
if (node.getAttribute("lang").getValue().equals("EN")){
infoHolder.englishTitle = value;
break;
}
}
node = node.getNext();
}
}
return infoHolder;
}
@Override
public void write(OutputNode node, InfoHolder value) throws Exception {
}
}
}
I also needed to instantiate a SimpleXmlConverterFactory
with a Serializer
using AnnotationStrategy
like so:
SimpleXmlConverterFactory factory = SimpleXmlConverterFactory.create(new Persister(new AnnotationStrategy()));
Using a custom Converter exposed the XML nodes, which allowed me to figure out if the info
node had img
children, and if not, get the node value itself.