Search code examples
javaandroidxml-parsingsaxsaxparser

Sax - ExpatParser$ParseException

I'm making an Android application that reads an XML Internet. This application uses SAX to parse XML. This is my code for the part of parsing:

public LectorSAX(String url){
    try{
        SAXParserFactory spf=SAXParserFactory.newInstance();
        SAXParser sp = spf.newSAXParser();
        DefaultHandler lxmlr=new LibraryXMLReader() ;
        sp.parse(url, lxmlr);

        nodo=((LibraryXMLReader)lxmlr).getNodoActual();

    }catch(ParserConfigurationException e){ 
        System.err.println("Error de parseo en LectorSAX.java: "+e);
    }catch(SAXException e){
        System.err.println("Error de sax LectorSAX.java: " + e);
    } catch (IOException e){
        System.err.println("Error de  io LectorSAX.java: " + e);
    }
}

The problem is that SAXException occurs. The exception message is as follows:

org.apache.harmony.xml.ExpatParser$ParseException: At line 4, column 42: not well-formed (invalid token)

However, if I put the same code in a normal Java SE application, this exception does not occur and everything works fine.

Why the same code works fine in a Java SE application, not an Android?. On the other hand, How to solve the problem?.

Thanks for the help.

Greetings.

Solution

  • This could be a character encoding problem.
    As you can see, the invalid token error points to the line #4.
    In this line, you can find an acute (Meteorología) and a tilde (España). The XML header shows a ISO-8859-15 encoding value. As it's less common than UTFs or ISO-8859-1 encodings, this could result in a error when the SAXParser connects and try to convert the byte content into chars using your system default charset.

    Then, you'll need to tell the SAXParser which charset to use. A way to do so, is to pass an InputSource, instead of the URL, to the parse method. As an example:

    SAXParserFactory factory = SAXParserFactory.newInstance();
SAXParser parser = factory.newSAXParser();

InputSource is = new InputSource(url);
is.setEncoding("ISO-8859-15");

DefaultHandler lxmlr=new LibraryXMLReader() ;
sp.parse(is, lxmlr);

    EDIT: It seems that Android VM does not support this encoding, throwing a org.apache.harmony.xml.ExpatParser$ParseException: At line 1, column 0: unknown encoding exception.
    As ISO-8859-15 it's mainly compatible with ISO-8859-1, except some specific characters (as you can see here), a workaround is changing the ISO-8859-15 value to ISO-8859-1 at the setEncoding method, forcing the parser to use a different but compatible charset encoding: 

    is.setEncoding("ISO-8859-1");

    As it seems, as Android doesn't support the declared charset, it uses its default (UTF-8) and hence the parser can't use the XML declaration to choose the apropiate encoding.