I am trying to read text from a web document using a BufferedReader over an InputStreamReader on an URL (to the file on some Apache server).
String result = "";
URL url = new URL("http://someserver.domain/somefile");
BufferedReader in = null;
in = new BufferedReader(new InputStreamReader(url.openStream(), "iso-8859-1"));
result += in.readLine();
Now this works just fine. But Obviously I'd like the reader not to just read one line, but as many as there are in the file.
Looking at the BufferedReader API the following code should do just that:
while (in.ready()) {
result += in.readLine();
}
I.e. read all lines while there are more lines, stop when no more lines are there. This code does not work however - the reader just never reports ready() = true!
I can even print the ready() value right before reading a line (which reads the correct string from the file) but the reader will report 'false'.
Am I doing something wrong? Why does the BufferedReader return 'false' on ready when there is actually stuff to read?
ready() != has more
ready()
does not indicate that there is more data to be read. It only shows if a read will could block the thread. It is likely that it will return false before you read all data.
To find out if there is no more data check if readLine()
returns null
.
String line = in.readLine();
while(line != null){
...
line = in.readLine();
}
Another way you can do this that bypasses the in.ready()
is something like:
while ((nextLine = in.readLine()) != null) {
result += nextLine;
}
You will just continue reading until you are done. This way you do not need to worry about the problem with in.ready()
.
I think the standard way to write this is to just attempt to read the line and verify that it returned sometime. Something like this:
while ((String nextLine = in.readLine()) != null) {
//System.out.println(nextLine);
result += nextLine;
}
So you just continue to go until you get null returned from the stream. See here for extra information:
http://download.oracle.com/javase/1.5.0/docs/api/java/io/BufferedReader.html#readLine()
The BufferedReader.ready()
method is behaving as specified:
The Reader.ready()
javadoc says the following:
[Returns] true
if the next read()
is guaranteed not to block for input, false
otherwise. Note that returning false
does not guarantee that the next read will block.
Then the BufferedReader.ready()
javadoc says the following:
Tells whether this stream is ready to be read. A buffered character stream is ready if the buffer is not empty, or if the underlying character stream is ready.
If you put these two together, it is clear that BufferedReader.ready()
can return false
in situations where are characters available. In short, you shouldn't rely on ready()
to test for logical end-of-file or end-of-stream.
This is what we have been using consistently for years - not sure if it is the "standard" method. I'd like to hear comments about the pros and cons of using URL.openURLStream() directly, and if that is causing the OP's problems. This code works for both HTTP and HTTPS connections.
URL getURL = new URL (servletURL.toString() + identifier+"?"+key+"="+value);
URLConnection uConn = getURL.openConnection();
BufferedReader br = new BufferedReader (new
InputStreamReader (uConn.getInputStream()));
for (String s = br.readLine() ; s != null ; s = br.readLine()) {
System.out.println ("[ServletOut] " + s);
// do stuff with s
}
br.close();
Basically the BufferedReader.ready() method can be used for checking whether the underlying stream is ready for providing data to the method caller.... else we can wait the thread for some time till it becomes ready.
But the real problem is that after we completely read the data stream, it will throw false..
so we didn't know whether the stream is fully read OR underlying stream is busy....
If you want to use in.ready(), the following worked for me well:
for (int i = 0; i < 10; i++) {
System.out.println("is InputStreamReader ready: " + in.ready());
if (!in.ready()) {
Thread.sleep(1000);
} else {
break;
}
}