The Java Native Interface (JNI) - JavaTech: An Introduction to Scientific and Technical Computing with Java - page 609

NoSuchFieldError can also result from passing the wrong jclass refer-
ence, resulting in a lookup of the correct field name and signature but in a class
that does not contain that field.
Like most everything else in JNI, type signatures can appear messy, but they
are well-defined and straightforward. It is easiest to introduce the type signature
notation by way of an example. Suppose we have a Java class named JNIDemo
like this:
package javatech.jni22;
class JNIDemo
{
static float a - static - float;
int some - int;
int[] array;
int[][] array2d;
String some - string;
MyCustomObject my - custom;
...
}
The type signatures follow a well-defined and deterministic formula. All the
Java primitive types have a single-letter designation - I for int , F for float ,
D for double , etc. So the first two fields are simple. The type signature of
a - static - float is “ F „ and for some - int , the type signature is “ I „ .
With this information, we can get the field IDs of the float and int fields
as follows:
jfieldID a - static - float - fid =
jenv->GetStaticFieldID (cls, "a - static - float", "F");
if (a - static - float - fid == NULL) return NULL;
and
jfieldID some - int - fid =
jenv->GetFieldID (cls, "some - int", "I");
if (some - int - fid == NULL) return NULL;
Arrays of primitives use the same single-letter mapping as the primitive prefixed
with a left square bracket. So the signature of the int[] array is “ [I „ .Youcan
think of this “ [I „ notation meaning “array of type I.” Thus,
jfieldID array - fid = jenv- > GetFieldID (cls, " array " , " [I " );
A float[] array would obviously be “ [F „ and the 2D int array ( int[][]
on the Java side) has a type signature of “ [[I „ .
The type signature of an object field is more complicated. It begins with the
letter L followed by the fully qualified name of the object type with slashes ( / )