Java Socket HTTP GET request

Question

I'm trying to create a simple Java program that create an HTTP request to a HTTP server hosted locally, by using Socket.

This is my code:

try
    {
        //Create Connection
        Socket s = new Socket("localhost",80);
        System.out.println("[CONNECTED]");
        DataOutputStream out = new DataOutputStream(s.getOutputStream());
        DataInputStream in   = new DataInputStream(s.getInputStream());


        String header = "GET / HTTP/1.1\n"
                +"Host:localhost\n\n";
        byte[] byteHeader = header.getBytes();
        out.write(byteHeader,0,header.length());

        String res = "";
        /////////////READ PROCESS/////////////
        byte[] buf = new byte[in.available()];
        in.readFully(buf);
        System.out.println("\t[READ PROCESS]");
        System.out.println("\t\tbuff length->"+buf.length);
        for(byte b : buf)
        {
            res += (char) b;
        }
        System.out.println("\t[/READ PROCESS]");
        /////////////END READ PROCESS/////////////

        System.out.println("[RES]");
        System.out.println(res);
        System.out.println("[CONN CLOSE]");

        in.close();
        out.close();
        s.close();


    }catch(Exception e)
    {
        e.printStackTrace();
    }

But by when I run it the Server reponse with a '400 Bad request error'. What is the problem? Maybe some HTTP headers to add but I don't know which one to add.

Is using sockets mandatory? Java provides other classes which may simplify your life: [How do I do a HTTP GET in Java?](https://stackoverflow.com/q/1485708) — Pshemo, Feb 16 '18 at 09:54
So many errors, so little time. No `Content-length` header. Wrong line terminator. Classic misuse of `available()`. Wrong order of closes. Why are you trying to write this for yourself when Java already does it for you? — user207421, Feb 16 '18 at 10:05

score 3 · Accepted Answer · answered Feb 16 '18 at 10:15

There are a couple of issues with your request:

String header = "GET / HTTP/1.1\n"
             + "Host:localhost\n\n";

The line break to be used must be Carriage-Return/Newline, i.e. you should change that to

String header = "GET / HTTP/1.1\r\n"
             + "Host:localhost\r\n\r\n";

Next problem comes when you write the data to the OutputStream:

    byte[] byteHeader = header.getBytes();
    out.write(byteHeader,0,header.length());

The call of readBytes without the specification of a charset uses the system's charset which might be a different than the one that is needed here, better use getBytes("8859_1"). When writing to the stream, you use header.length() which might be different from the length of the resulting byte-array if the charset being used leads to the conversion of one character into multiple bytes (e.g. with UTF-8 as encoding). Better use byteHeader.length.

    out.write(byteHeader,0,header.length());

    String res = "";
    /////////////READ PROCESS/////////////
    byte[] buf = new byte[in.available()];

After sending the header data you should do a flush on the OutputStream to make sure that no internal buffer in the streams being used prevents the data to actually be sent to the server.

in.available() only returns the number of bytes you can read from the InputStream without blocking. It's not the length of the data being returned from the server. As a simple solution for starters, you can add Connection: close\r\n to your header data and simply read the data you're receiving from the server until it closes the connection:

StringBuffer sb = new StringBuffer();
byte[] buf = new byte[4096];
int read;
while ((read = in.read(buf)) != -1) {
    sb.append(new String(buf, 0, read, "8859_1"));
}
String res = sb.toString();

Oh and independent form the topic of doing an HTTP request by your own:

    String res = "";
    for(byte b : buf)
    {
        res += (char) b;
    }

This is a performance and memory nightmare because Java is actually caching all strings in memory in order to reuse them. So the internal cache gets filled with each result of this concatenation. A response of 100 KB size would mean that at least 5 GB of memory are allocated during that time leading to a lot of garbage collection runs in the process.

Oh, and about the response of the server: This most likely comes from the invalid line breaks being used. The server will regard the whole header including the empty line as a single line and complains about the wrong format of the GET-request due to additional data after the HTTP/1.1.

'Java is actually caching all strings in memory in order to reuse them': explain. Justify. Cite. — user207421, Feb 16 '18 at 10:37

score 1 · Answer 2 · edited Oct 07 '21 at 11:21

1

According to HTTP 1.1:

HTTP/1.1 defines the sequence CR LF as the end-of-line marker for all protocol elements except the entity-body [...].

So, you'll need all of your request to be ending with \r\n.

edited Oct 07 '21 at 11:21

Community

1
1

answered Feb 16 '18 at 10:01

Little Santi

8,563
2
18
46

Java Socket HTTP GET request

2 Answers2