LibJS: Change an error message used by %TypedArray%.prototype.with

ErrorType::InvalidIndex does not encapsulate the reasons why an index
may be invalid. For example:

    let array = new Uint8Array([1, 2, 3, 4, 5]);
    array.with(10, 0);

Will currently yield:

    [RangeError] Index must be a positive integer

Which is misleading because 10 *is* a positive integer.

Userland/Libraries/LibJS/Runtime/ErrorTypes.h

@@ -296,6 +296,7 @@
     M(TypedArrayInvalidBufferLength, "Invalid buffer length for {}: must be a multiple of {}, got {}")                                  \
     M(TypedArrayInvalidByteOffset, "Invalid byte offset for {}: must be a multiple of {}, got {}")                                      \
     M(TypedArrayInvalidCopy, "Copy between arrays of different content types ({} and {}) is prohibited")                                \
+    M(TypedArrayInvalidIntegerIndex, "Invalid integer index: {}")                                                                       \
     M(TypedArrayInvalidTargetOffset, "Invalid target offset: must be {}")                                                               \
     M(TypedArrayOutOfRangeByteOffset, "Typed array byte offset {} is out of range for buffer with length {}")                           \
     M(TypedArrayOutOfRangeByteOffsetOrLength, "Typed array range {}:{} is out of range for buffer with length {}")                      \

Userland/Libraries/LibJS/Runtime/TypedArrayPrototype.cpp

@@ -1680,7 +1680,7 @@ JS_DEFINE_NATIVE_FUNCTION(TypedArrayPrototype::with)
 
     // 9. If ! IsValidIntegerIndex(O, 𝔽(actualIndex)) is false, throw a RangeError exception.
     if (!is_valid_integer_index(*typed_array, CanonicalIndex(CanonicalIndex::Type::Index, actual_index)))
-        return vm.throw_completion<RangeError>(ErrorType::InvalidIndex);
+        return vm.throw_completion<RangeError>(ErrorType::TypedArrayInvalidIntegerIndex, actual_index);
 
     // 10. Let A be ? TypedArrayCreateSameType(O, « 𝔽(len) »).
     MarkedVector<Value> arguments(vm.heap());