Question

I am working on a simple XML parser in Java. I have been using DocumentBuilderFactory successfully with many sources, but one of my new ones is a collection of individual nodes.

The file.xml looks like this:

<XML Version....>
<!DOCTYPE...>
<main_document_node>
.....others....
</main_document_node>
<XML Version....>
<!DOCTYPE...>
<main_document_node>
.....others....
 </main_document_node>

I had been using a command like this:

DocumentBuilderFactory dbf = DocumentBuilderFactory.newInstance();
Document d = dbf.newDocumentBuilder().parse(/path/file.xml);

But that doesn’t seem to work with multiple parent nodes. Is there an easier way than making a temporary working file that I write a main_document_node to? That pseudo-code would be

new Writer temp.xml
new Reader file.xml
while not at the end of file.xml
    read/write first main_document_node to temp
    parse temp.xml

I think there should be a way to use the inputsource/stream option of the DocumentBuilderFactory, but I’m not sure how to go about doing that.

Thanks!

Answer 1

You could skip the write to file step by creating an InputStream from the computet string fragment (for one node):

new Reader file.xml
while not at the end of file.xml {
    String node = read_first_main_document_node();
    InputStream is = new ByteArrayInputStream( node.getBytes( charset ) );
    parse(is); 
}