来自HttpURLConnection的对象的InputStream无法获取所有数据

Question

So I have an HttpURLConnection connecting to some image on some URL. I am downloading this image using the Range property (downloading one file in separate parallel connections). After I grab all of the parts, I stitch them back together. Everything works great unless I try to do an InputStream with a larger file (meaning that I can get a file in 200 parallel connections, but not 5).

For example:

Say I want to download the object: http://stuorgs.uga.edu/images/spotlight/spotlight_button.png in just 1 part. So my range is from bytes=0-max. My program creates a byte array of size max, and the InputStream from the HttpURLConnection is read into the array. However, the InputStream, after a certain point, either only sends over 0000, or closes prematurely.

Here is my code:

try {
    HttpURLConnection urlConnection = (HttpURLConnection)url.openConnection();
    String bytesToGrab = "bytes=" + startingByte + "-" + endingByte;
    urlConnection.setRequestProperty("Range", bytesToGrab);

    int sizeOfData = endingByte - startingByte;
    byte[] storeData = new byte[sizeOfData];

    InputStream inputStream = urlConnection.getInputStream();
    inputStream.read(storeData);
    inputStream.close();

    createFile(storeData); // this takes the byte array and creates a file with it
} catch (IOException e) {
    System.out.println("Cannot open url connection: " + e);
} 

So, is my InputStream closing early? If so, how do I prevent that from happening? Otherwise, what is wrong?

Answer 1

Nowhere in the Javadoc does it state that InputStreamread(byte[]) fills the buffer. It blocks until at least one byte of data is available and then transfers whatever is available, and returns a count of bytes transferred.

You have to loop. Or use DataInputStream.readFully() .

NB Don't name your variables after classes, and don't start the names with capital letters. That's for classes.