Is it possible to check progress of URLconnection.getInputStream()?

latata picture latata · Mar 2, 2011 · Viewed 8.2k times · Source

I want to check progress of downloading file by URLconnection. Is it possible or should I use another library? This is my urlconnection function:

public static String sendPostRequest(String httpURL, String data) throws UnsupportedEncodingException, MalformedURLException, IOException {
    URL url = new URL(httpURL);

    URLConnection conn = url.openConnection();
    //conn.addRequestProperty("Content-Type", "text/html; charset=iso-8859-2");
    conn.setDoOutput(true);
    OutputStreamWriter wr = new OutputStreamWriter(conn.getOutputStream());
    wr.write(data);
    wr.flush();

    BufferedReader rd = new BufferedReader(new InputStreamReader(conn.getInputStream(), "ISO-8859-2"));
    String line, all = "";
    while ((line = rd.readLine()) != null) {
        all = all + line;
    }
    wr.close();
    rd.close();
    return all;
}

I understand that whole file is downloaded in this line (or worng)?:

BufferedReader rd = new BufferedReader(new InputStreamReader(conn.getInputStream(), "ISO-8859-2"));

So is it possible to do this in this code?

Answer

BalusC picture BalusC · Mar 2, 2011

Just check if the HTTP Content-Length header is present in the response.

int contentLength = connection.getContentLength();

if (contentLength != -1) {
    // Just do (readBytes / contentLength) * 100 to calculate the percentage.
} else {
    // You're lost. Show "Progress: unknown"
}

Update as per your update, you're wrapping the InputStream inside a BufferedReader and reading inside a while loop. You can count the bytes as follows:

int readBytes = 0;

while ((line = rd.readLine()) != null) {
    readBytes += line.getBytes("ISO-8859-2").length + 2; // CRLF bytes!!
    // Do something with line.
}

The + 2 is to cover the CRLF (carriage return and linefeed) bytes which are eaten by BufferedReader#readLine(). More clean approach would be to just read it by InputStream#read(buffer) so that you don't need to massage the bytes forth and back from characters to calculate the read bytes.

See also: