Issue with Translation API in Java

Question

i am trying to translate from english into Arabic using translate.googleapis.com.

it works well with all letters except one letter it always show letter 'ف' as '�?' any suggestions ?

private static String callUrlAndParseResult(String langFrom, String langTo, String word) throws Exception {
    String url =
        "https://translate.googleapis.com/translate_a/single?" + "client=gtx&" + "sl=" + langFrom + "&tl=" +
        langTo + "&dt=t&q=" + URLEncoder.encode(word, "UTF-8");
    URL obj = new URL(url);
    URLConnection con = obj.openConnection();
    con.setRequestProperty("User-Agent", "Mozilla/5.0");

    BufferedReader in = new BufferedReader(new InputStreamReader(con.getInputStream()));
    String inputLine;
    StringBuffer response = new StringBuffer();

    while ((inputLine = in.readLine()) != null) {
        response.append(inputLine);
    }
    in.close();

    return parseResult(response.toString());
}

private static String parseResult(String inputJson) throws Exception {
    /*
   * inputJson for word 'hello' translated to language Hindi from English-
   * [[["??????","hello",,,1]],,"en"]
   * We have to get '?????? ' from this json.
   */

    JSONArray jsonArray = new JSONArray(inputJson);
    JSONArray jsonArray2 = (JSONArray) jsonArray.get(0);
    JSONArray jsonArray3 = (JSONArray) jsonArray2.get(0);

    return jsonArray3.get(0).toString();
}

public static void main(String[] args) {
    try {
        String word = callUrlAndParseResult("en", "ar", "phone");
        System.out.println(new String(word.getBytes(), Charset.forName("UTF-8")));
    } catch (Exception e) {
        System.out.println(e.getMessage());
    }
}

i am using jdeveloper 12cR2

Answer 1

Please note that whenever you use Reader, there will be conversions between charsets. If you do not specify your charset, it will use the system default charset to encode the incoming byte stream, and you would get into trouble if the incoming byte stream is actually not in the same charset with your system.

Therefore, it is advised to specific the charset when using Reader.

So your code should like below.

private static String callUrlAndParseResult(String langFrom, String langTo, String word) throws Exception {
    String url =
        "https://translate.googleapis.com/translate_a/single?" + "client=gtx&" + "sl=" + langFrom + "&tl=" +
        langTo + "&dt=t&q=" + URLEncoder.encode(word, "UTF-8");
    URL obj = new URL(url);
    URLConnection con = obj.openConnection();
    con.setRequestProperty("User-Agent", "Mozilla/5.0");

    BufferedReader in = new BufferedReader(new InputStreamReader(con.getInputStream(), "UTF-8"));
    String inputLine;
    StringBuffer response = new StringBuffer();

    while ((inputLine = in.readLine()) != null) {
        response.append(inputLine);
    }
    in.close();

    return parseResult(response.toString());
}

private static String parseResult(String inputJson) throws Exception {
    /*
   * inputJson for word 'hello' translated to language Hindi from English-
   * [[["??????","hello",,,1]],,"en"]
   * We have to get '?????? ' from this json.
   */

    JSONArray jsonArray = new JSONArray(inputJson);
    JSONArray jsonArray2 = (JSONArray) jsonArray.get(0);
    JSONArray jsonArray3 = (JSONArray) jsonArray2.get(0);

    return jsonArray3.get(0).toString();
}

public static void main(String[] args) {
    try {
        String word = callUrlAndParseResult("en", "ar", "phone");
        System.out.println(word);
    } catch (Exception e) {
        System.out.println(e.getMessage());
    }
}