how to skip xml elements while loading into pojo (in xstream/jaxb)

I have a very big and complex xml and i want to load only selected fields into my object. I have tried with xstream but what i understood is my xml structure must be similar to my pojo. I am providing sample i/p and o/p for better understanding

<note>
<to>Tove</to>
<from>Jani</from>
<heading>Reminder</heading>
<body>Don't forget me this weekend!</body>
</note>

Pojo will be

    class Note{
     long noteId;
     String body;

    //getters and setters

   }

Question is how to skip xml elements while loading it into pojo?

Solution

You can do it with Jackson XML librairy. Jackson is maybe more famous for JSON serialising/deserialising but it has a XML extension.

If your XML file is big, look at the above link to see how you can partially/incrementally read it in order to get the best performances.

Anyway following is an example on how to read only a subset of properties using the @JsonIgnore annotation on the fields you don't want deserialised:

@JacksonXmlRootElement(localName = "jokes")
class Jokes {

    @JacksonXmlProperty(localName = "joke")
    @JacksonXmlElementWrapper(useWrapping = false)
    private Joke[] jokes;

    // getter, setter omitted
}

class Joke {
    @JacksonXmlProperty(isAttribute = true)
    long id;

    @JacksonXmlProperty(localName = "title")
    String title;

    @JsonIgnore
    String author;

    @JacksonXmlProperty(localName = "like")
    long like;

    @JsonIgnore
    String content;

    public String toString() {
        return Arrays.stream(new Object[] {id, title, author, like, content})
                .map(o -> o!=null ? o.toString() : "EMPTY")
                .collect(Collectors.joining(", ", "[", "]"));
    }

    // getters, setters omitted
}

With a sample file:

<jokes>
    <joke id="123">
        <title>C'est l'histoire d'un mec</title>
        <author>Coluche</author>
        <content>Blah blah blah</content>
        <like>4509</like>
    </joke>
    <joke id="777">
        <title>Si j'ai bien tout lu freud</title>
        <author>Coluche</author>
        <content>Blah blah blah</content>
        <like>345</like>
    </joke>
</jokes>

And the main()

public class XmlJokeReader {

    public static void main(String[] args) throws JsonProcessingException, IOException {
        XmlMapper mapper = new XmlMapper();

        Jokes jokesDoc = mapper.readValue(new File("./data/jokes.xml"), Jokes.class);
        Arrays.stream(jokesDoc.getJokes()).forEach(j -> System.out.println(j.toString()));
    }
}

The output is (note the EMPTY fields):

[123, C'est l'histoire d'un mec, EMPTY, 4509, EMPTY]
[777, Si j'ai bien tout lu freud, EMPTY, 345, EMPTY]

You can also create a pojo that contains only the fields you need - then not using @JsonIgnore. For that the XmlMapper has to be informed to ignore unknown XML properties.

    XmlMapper mapper = new XmlMapper();
    mapper.configure(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES, false);

EDIT

The full example for the case where you want a pojo with only a few fields:

Let say we have a pojo with only id and title:

class Joke {
    @JacksonXmlProperty(isAttribute = true)
    long id;

    @JacksonXmlProperty(localName = "title")
    String title;
    
    public String toString() {
        return new StringBuffer().append('[')
                .append(id).append(',')
                .append(title).append(']').toString();
    }
    // getters setters 
}

Executing the following main() with the xml file described above:

public class XmlJokeReader {

    public static void main(String[] args) throws JsonProcessingException, IOException {
        XmlMapper mapper = new XmlMapper();
        mapper.configure(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES, false);

        Jokes jokesDoc = mapper.readValue(new File("./data/jokes.xml"), Jokes.class);
        for (Joke joke : jokesDoc.getJokes()) {
           System.out.println(joke.toString());
        }
    }
}

Will give:

[123,C'est l'histoire d'un mec]
[777,Si j'ai bien tout lu freud]