How to remove � from a String?

Question

What is that char �? and how to remove it from a String? I got it from a BufferedReader and i got it because i read the contents in a char array and this array has to be assigned to a particular size.So, i got the String like that "aaaaaaa��", and I tried trim and subString but didn't change anything:

 String a = "aaaaaaa����";
//subString
    int i = a.lastIndexOf("a");
    a = a.substring(0, i+1);
//trim
    a = a.trim();

And this is my way to read the input:

BufferedReader in = new BufferedReader(new InputStreamReader(socket.getInputStream()));
char[] a = new char[1000];
int line;
String responseLine, server_response = "";
while((line = in.read(a)) != -1) {
      responseLine = String.valueOf(a);
      server_response = server_response + responseLine;
     }
in.close();
return server_response;

@donfuxx I have doubts in that too, but I don't know how to handle it — MRefaat, Mar 09 '14 at 13:51
@Wooble I'm pretty sure that this is not related to the data as I'm already now what's the data — MRefaat, Mar 09 '14 at 13:54
You open a `Reader` without specifying the encoding; as such the default JRE ecoding will be used. Is that what you want? — fge, Mar 09 '14 at 14:21
No, as the problem not comes from reading the input string, it comes from converting the non-filled char_array to a String — MRefaat, Mar 09 '14 at 14:23
Sorry to contradict you, but that may very well be the problem. Don't forget that a `Reader` takes the bytes of the stream and converts them to chars _depending on the encoding_. You never send `char`s down the wire, only bytes. The fact that you have bizarre characters showing up in your resulting strings is a sign that you don't use the correct encoding. Ultimately, a `String` is an array of `char`s, not bytes — fge, Mar 09 '14 at 14:25
Notably, a Java `char` is a very different thing from a C `char`. — chrylis -cautiouslyoptimistic-, Mar 09 '14 at 14:36

score 2 · Answer 1 · answered Mar 09 '14 at 14:31

This is very likely to be an encoding problem; you do not specify the encoding on your InputStreamReader, as such the system default is used.

Try and use:

new InputStreamReader(socket.getInputStream(), StandardCharsets.UTF_8)

instead.

If you are still stuck with JDK 6, replace StandardCharsets.UTF_8 with Charset.forName("UTF-8").

If you are unsure what encoding is used at the other end, you should not use a Reader but read the contents into a byte array. Then you can use a CharsetDecoder to try and map the bytes read into one or more encodings.

Example:

StandardCharsets.ASCII.newDecoder()

What didn't solve the problem? Have you tried the `CharsetDecoder` way? Do you at least know the encoding used by the other end? — fge, Mar 09 '14 at 14:53

score 1 · Answer 2 · answered Mar 09 '14 at 14:03

1

Try with unicode

Unicode corresponding to � is \ufffd

String str0 = "aaaaaaa����";
System.out.println(str0.replaceAll("\ufffd", ""));

answered Mar 09 '14 at 14:03

Rakesh KR

6,357
5
40
55

it didn't do anything also – MRefaat Mar 09 '14 at 14:10
Except that that display character is often used for "any character I don't have a glyph for". – chrylis -cautiouslyoptimistic- Mar 09 '14 at 14:22
@chrylis yea it may be that, do you have an idea to overcome that? – MRefaat Mar 09 '14 at 14:25

MRefaat · Accepted Answer · 2014-03-18T07:51:25.117

finally i found a way to solve that, it's not a professional one but efficient enough. all i had to do is filling the char array with white spaces just before starting the while loop and then after receiving the whole response i have just to trim it before returning it :

BufferedReader in = new BufferedReader(new InputStreamReader(socket.getInputStream()));
char[] a = new char[1000];
int line;
String responseLine, server_response = "";
for(int i = 0; i < a.length; i++){ //
      a[i] = ' ';                  // this is the for loop i added
    }                              //
while((line = in.read(a)) != -1) {
      responseLine = String.valueOf(a);
      server_response = server_response + responseLine;
      for(int i = 0; i < a.length; i++){ //
          a[i] = ' ';                    // this is the for loop i added
        }                                //
     }
in.close();
return server_response.trim();     // this is where i return the response trimmed

score 0 · Answer 4 · answered Mar 09 '14 at 13:56

0

you could handle it like this:

System.out.println("aaaaaaa����".replace("�", ""));

remaining string will be aaaaaaa.

I recommend to investigate the input source though and figure out why you get those chars there. Probably there is somewhere an issue with the encoding.

answered Mar 09 '14 at 13:56

donfuxx

11,277
6
44
76

I tried that, it didn't agree to save and tells me `some characters cannot be mapped using "Cp1252"` – MRefaat Mar 09 '14 at 14:07
1

seems like this encoding is not supported. Try using utf-8 encoding in your reader like suggested here: http://stackoverflow.com/a/4964652/2399024 – donfuxx Mar 09 '14 at 14:20
the problem not comes from reading the input string, it comes from converting the non-filled char_array to a String – MRefaat Mar 09 '14 at 14:24

score 0 · Answer 5 · answered Mar 09 '14 at 13:56

0

If your only expecting numbers and characters you can run a for loop on the byte array and run Char.isLetterOrDigit on each character replacing those that arent with ""

answered Mar 09 '14 at 13:56

user3293629

119
1
8

there are some special chars – MRefaat Mar 09 '14 at 14:05
This will return false if the character isn't alphanumeric So you can use this to find the special chars and set them to "" – user3293629 Mar 09 '14 at 14:10
there are some special chars that I need to be existed not to be removed – MRefaat Mar 09 '14 at 14:11
If that's the only character you need removed you can try to compare each char to that character and replace – user3293629 Mar 09 '14 at 14:23
I can't write that character into my code in order to compare it, it gives me `some characters cannot be mapped using "Cp1252"` – MRefaat Mar 09 '14 at 14:42
Do you know specifically what special character you want? – user3293629 Mar 09 '14 at 15:23
1

@MRefaat _it gives me some characters cannot be mapped using "Cp1252"_: are you using Eclipse IDE? Try setting your workspace text file encoding to UTF-8. You can find it in the Preferences Menu: Window -> Preferences -> General -> Workspace, then select "Other: UTF-8" as "Text file encoding" – Modus Tollens Mar 10 '14 at 14:39
@KatjaChristiansen that really enabled me to read the char but it didn't replaced from the output when i made some replacing processes – MRefaat Mar 11 '14 at 07:32

How to remove � from a String?

5 Answers5