29

In my java project, I'm passing FileInputStream to a function, I need to convert (typecast FileInputStream to string), How to do it.??

public static void checkfor(FileInputStream fis) {
   String a=new String;
   a=fis         //how to do convert fileInputStream into string
   print string here
}
Aubin
  • 14,617
  • 9
  • 61
  • 84
bhushan23
  • 481
  • 1
  • 4
  • 13

6 Answers6

30

You can't directly convert it to string. You should implement something like this Add this code to your method

    //Commented this out because this is not the efficient way to achieve that
    //StringBuilder builder = new StringBuilder();
    //int ch;
    //while((ch = fis.read()) != -1){
    //  builder.append((char)ch);
    //}
    //          
    //System.out.println(builder.toString());

Use Aubin's solution:

public static String getFileContent(
   FileInputStream fis,
   String          encoding ) throws IOException
 {
   try( BufferedReader br =
           new BufferedReader( new InputStreamReader(fis, encoding )))
   {
      StringBuilder sb = new StringBuilder();
      String line;
      while(( line = br.readLine()) != null ) {
         sb.append( line );
         sb.append( '\n' );
      }
      return sb.toString();
   }
}
Arsen Alexanyan
  • 3,061
  • 5
  • 25
  • 45
  • 1
    May be :) I think he is not looking for the efficient way, he is just trying to learn – Arsen Alexanyan Mar 01 '13 at 15:59
  • Actually, reading characters can be more efficient that reading lines which are strings, because a to make a String, you have top copy each of the characters in to the String, and that is only to return a String which is then copied to another String.... – AgilePro Mar 01 '13 at 16:01
  • @Aubin Why it is not efficient? – Arsen Alexanyan Mar 01 '13 at 16:05
  • 1
    The native methods for moving a block of byte or characters can be faster than a loop that handles each character. However, what is more important is to look at whether there is needless extra search, converting, and copying of data, which accounts for more overhead than the character loop overhead. By the way, this answer is putting a byte into a character, which is a problem. – AgilePro Mar 01 '13 at 16:13
  • @AgilePro Why it is a problem converting byte into character? – Arsen Alexanyan Mar 01 '13 at 16:18
  • There is not a one-to-one relationship between bytes and characters. Sometimes, two or more bytes are needed to make one character depending upon charset. Thus you need to use a InputStreamReader to assure that the right number of bytes are read to compose a single character. – AgilePro Mar 01 '13 at 16:19
  • Ok, will note this. I was previously used this type of code many times but I haven't any problems yet. – Arsen Alexanyan Mar 01 '13 at 16:22
  • whew! problem solved at last :) i was afraid of reading input stream character by character and was fond of BufferedReader/FileReader and was trying to get absolute path from Uri... but atlast, this simple solution worked butterly :D – Hasaan Ali May 30 '16 at 18:46
24
public static String getFileContent(
   FileInputStream fis,
   String          encoding ) throws IOException
 {
   try( BufferedReader br =
           new BufferedReader( new InputStreamReader(fis, encoding )))
   {
      StringBuilder sb = new StringBuilder();
      String line;
      while(( line = br.readLine()) != null ) {
         sb.append( line );
         sb.append( '\n' );
      }
      return sb.toString();
   }
}
Aubin
  • 14,617
  • 9
  • 61
  • 84
  • Depending on the requirements of the Original Poster, perhaps needed the file encoding and line separator (or read as it appears in the file). – Paul Vargas Mar 01 '13 at 16:03
  • Yes, you're right but when reading stream to RAM we may accept the line terminator will be normalized into \n (assumption to debate) – Aubin Mar 01 '13 at 16:06
  • The close should happen in a finally block, or better in a try-with-resource block. – Puce Mar 01 '13 at 16:06
  • 2
    You need to specify the character encoding. You can not assume that the platform default encoding will be correct. – AgilePro Mar 01 '13 at 16:10
  • I'm using file as a parameter to this function and if a condition satisfies then i have to print whole path of a file that's why i took fileinputstream as a parameter.... – bhushan23 Mar 07 '13 at 15:31
16

Using Apache commons IOUtils function

import org.apache.commons.io.IOUtils;

InputStream inStream = new FileInputStream("filename.txt");
String body = IOUtils.toString(inStream, StandardCharsets.UTF_8.name());
atr
  • 684
  • 6
  • 10
3

Don't make the mistake of relying upon or needlessly converting/losing endline characters. Do it character by character. Don't forget to use the proper character encoding to interpres the stream.

public String getFileContent( FileInputStream fis ) {
    StringBuilder sb = new StringBuilder();
    Reader r = new InputStreamReader(fis, "UTF-8");  //or whatever encoding
    int ch = r.read();
    while(ch >= 0) {
        sb.append(ch);
        ch = r.read();
    }
    return sb.toString();
}

If you want to make this a little more efficient, you can use arrays of characters instead, but to be honest, looping over the characters can be still quite fast.

public String getFileContent( FileInputStream fis ) {
    StringBuilder sb = new StringBuilder();
    Reader r = new InputStreamReader(fis, "UTF-8");  //or whatever encoding
    char[] buf = new char[1024];
    int amt = r.read(buf);
    while(amt > 0) {
        sb.append(buf, 0, amt);
        amt = r.read(buf);
    }
    return sb.toString();
}
AgilePro
  • 5,588
  • 4
  • 33
  • 56
  • Actually, there is good reason to believe it is more efficient than the code that is scanning for, and manipulating, the end of line characters, then copying that into a string for return. Character by character efficiently works with the native content of the stream. – AgilePro Mar 01 '13 at 15:57
  • sorry, forgot the Reader to convert bytes to chars. – AgilePro Mar 01 '13 at 16:03
  • 2
    You can not declare a variable in the while expression part. And the read of a Reader returns int, not char. – Roger Lindsjö Mar 01 '13 at 16:07
  • Yes, I actually prefer to keep the declaration out of the block condition, even though it means you have two lines with read statements. I have changed it thus. – AgilePro Mar 01 '13 at 16:15
1

From an answer I edited here:

static String convertStreamToString(java.io.InputStream is) {
    if (is == null) {
        return "";
    }

    java.util.Scanner s = new java.util.Scanner(is);
    s.useDelimiter("\\A");

    String streamString = s.hasNext() ? s.next() : "";

    s.close();

    return streamString;
}

This avoids all errors and works well.

Andrew
  • 5,839
  • 1
  • 51
  • 72
1

Use following code ---->

try {
    FileInputStream fis=new FileInputStream("filename.txt");
    int i=0;    
    while((i = fis.read()) !=-1 ) {  // to reach until the laste bytecode -1
        System.out.print((char)i);   /* For converting each bytecode into character */ 
    }
    fis.close();
} catch(Exception ex) {
    System.out.println(ex); 
}
Thomas Fritsch
  • 9,639
  • 33
  • 37
  • 49
Mohammad Zuha Khalid
  • 87
  • 9