Question

I have a simple code that gets xml file from given URL:

DocumentBuilderFactory.newInstance().newDocumentBuilder().parse(link);

that code returns xml document (org.w3c.dom.Document). I just need to get size of resulting xml document. Is there any elegant way to do it, WITHOUT involving third-party jars?

P.S. size in KB, or MB, not number of nods

Answer 1

First naive version: Load the file into a local buffer. Then you know how long is your input. Then parse the XML out of the buffer:

URL url = new URL("...");
InputStream in = new BufferedInputStream(url.openStream());
ByteArrayOutputStream buffer1 = new ByteArrayOutputStream();
int c = 0;
while((c = in.read()) >= 0) {
  buffer1.write(c);
}

System.out.println(String.format("Length in Bytes: %d", 
    buffer1.toByteArray().length));

ByteArrayInputStream buffer2 = new ByteArrayInputStream(buffer1.toByteArray());

Document doc = DocumentBuilderFactory.newInstance()
    .newDocumentBuilder().parse(buffer2);

Drawback is the additional buffer in RAM.

Second more elegant version: Wrap the input stream with a custom java.io.FilterInputStream counting the bytes streaming through it:

URL url = new URL("...");
CountInputStream in = new CountInputStream(url.openStream());
Document doc = DocumentBuilderFactory.newInstance().newDocumentBuilder().parse(in);
System.out.println(String.format("Bytes: %d", in.getCount()));

Here is the CountInputStream. All read() methods are overwritten to delegate to the super class and count the resulting bytes:

public class CountInputStream extends FilterInputStream {

  private long count = 0L;

  public CountInputStream(InputStream in) {
    super(in);
  }

  public int read() throws IOException {
    final int c = super.read();
    if(c >= 0) {
      count++;
    }
    return c;
  }

  public int read(byte[] b, int off, int len) throws IOException {
    final int bytesRead = super.read(b, off, len);
    if(bytesRead > 0) {
      count += bytesRead;
    }
    return bytesRead;
  }

  public int read(byte[] b) throws IOException {
    final int bytesRead = super.read(b);
    if(bytesRead > 0) {
      count += bytesRead;
    }
    return bytesRead;
  }

  public long getCount() {
    return count;
  }
}