Difference between Text and String DataTypes in Hadoop

Here I am going to discuss some of the differences between Text and String class in Hadoop.

Text class lies in the package: import org.apache.hadoop.io.*;

Difference 1:
 
Text is not immutable :   String is immutable

Text t = new Text("hadoop");
t.set("BigData")
print "t" --> prints "BigData"


Difference 2 :
Text stores the string in a byte buffer with UTF-8 unicode encoding

Example : Text t = new Tex("hadoop");

will get converted into byte[] array, and then places in to ByteBuffer.

so the string "hadoop" will get stored like this [UTF-CODE(h),UTF-code(a)........ UTF-code(p)]

so this this is the byte[] array representation for string "hadoop"
   [104,97,100,111,111,112]

Why ?

Text uses standard UTF-8 which makes it potentially easier to inter-operate with other tools that
understand UTF-8.


Difference 3 :
CharAt(int index) in string returns the char at specified index.

charAt(int index) in Text returns the Unicode point in the above case it i 100.

Difference 4 : Due to lack of Rich API for manipulating strings in Text many cases we use to
convert it to String.

Difference 5 : Iterating over Text characters is tedious process when compared to string.
Example of iterating over charactes in Text;

Text t = new Text("hadoop");
      ByteBuffer bf = ByteBuffer.wrap(t.getBytes(),0,t.getLength());
 int cp;
 while(bf.hasRemaining()){
  cp = Text.bytesToCodePoint(bf);  
 System.out.print((char) cp);
 }

Similarity 1 :  find in Text equals to indexOf in String.

Text t = new Text("hadoop");
String s = new String("hadoop");

 System.out.println(" >>> "+t.find("o"));
 System.out.println(" >>> "+s.indexOf("o"));