The result of that function is another C pointer, which we then need to convert to a back into a String. Next, we read the string in from the JNIEnv and convert it into a C pointer to pass to rust_greeting. The JNIEnv will be the object we will use to read values associated with the pointers that we are taking as argument. extern defines the function as one that will be exposed to other languages.Īs arguments, along with the JString that our Java function declaration said that we will be providing, we also need to take an instance of the JNIEnv and a class reference (which is unused in this example). By marking the function as not memory safe, we are alerting other Rust functions that it may not be able to deal with a null pointer. This situation would never happen in Rust only code as the Rust compiler enforces memory safety. This function needs to be marked unsafe because we will be dealing with pointers from a language that allows null pointers, but our code doesn’t check for NULL. This flag will apply to all functions and variables created inside this module that we are creating, called android.Īfter declaring that we need the jni crate, and importing some useful objects from it, we can declare our function. Given the way that JNI constructs native function names, we need to tell the Rust compiler to go easy on us in this instance. However, we defined our class name and native method in our Android project using Java coding conventions which is camelCase and UpperCamelCase and we don’t want to change this or our Java code will look wrong. The Rust compiler is very strict - this is one of the things that makes Rust great - and it enforces the use of snake_case throughout. The second line, #, tells the compiler not to warn if we are not using snake_case for a variable or function name. # is a special attribute that allows you to compile code based on a flag passed to the compiler. The first line here # is telling the compiler to target Android when compiling this module.

Mozilla firefox for android x86