This is the first time that I use the JNI and also the first time that I have to write some lines in C.
What I am trying to do is very simple. I'm just trying to switch the endiannes of a byte[] using a C routine.
In java it is done like this:
public void switchEndianness(byte[] array){
byte byte1;
byte byte2;
for(int i = 0; i < array.length ; i+=2){
byte1 = array[i];
byte2 = array[i+1];
array[i] = byte2;
array[i+1] = byte1;
}
}
So to do this using JNI, I've tried to imlpement the same routine in the JNICALL, but it doesn't compile. What I've written so far is this:
JNIEXPORT void JNICALL Java_CEndianness_switchEndianness(JNIEnv *env, jobject obj, jbyteArray array, jint offset, jint length){
char byte1;
char byte2;
int i;
for(i = offset; i < length ; i+=2){
byte1 = array[i];
byte2 = array[i+1];
array[i] = byte2;
array[i+1] = byte1;
}
}
I have no clue how to use the jbyteArray type of data. is it possible to store a jbyte in a char??? Another question is.. when this routine is over...will the byte[] in java be modified?? Or is it only modified inside the C call?
Any help???
Thanks to everybody!
you can get jbyte* by GetByteArrayElements:
jbyte* bufferPtr = (*env)->GetByteArrayElements(env, array, NULL);
And it is important to know the length of your array:
jsize lengthOfArray = (*env)->GetArrayLength(env, array);
Having jbyte* and length, you can do all the things in c-array. Finally, releasing it:
(*env)->ReleaseByteArrayElements(env, array, bufferPtr, 0);