# US 2015/0106796 A1 · Class loading device for a java runtime environment, cluster system and method of executing a function

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2015/0106796 A1
- Document ID: US20150106796A1
- Status: Published application
- Kind: A1
- Application number: 14/513,651
- Filed: 2014-10-14
- Published: 2015-04-16
- Earliest claimed: 2013-10-15
- Assignees: Fujitsu Technology Solutions Intellectual Property GmbH
- Applicants: Fujitsu Technology Solutions Intellectual Property GmbH
- Inventors: Michael Schuler
- Web page: https://patentgrep.com/patent/US20150106796A1

## CPC

- G06F 9/44521: {Dynamic linking or loading; Link editing at or after load time, e.g. Java class loading}
- G06F 8/41: Compilation
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A class loading device for a Java runtime environment, wherein the class loading device is configured to load predetermined source code written in R programming language to parse the R source code and, based on the parsed R source code, generate corresponding bytecode at least of a first Java class and transfer same to the Java runtime environment.

## Claims

### Claim 1

A class loading device for a Java runtime environment, wherein the class loading device is configured to load predetermined source code written in R programming language to parse the R source code and, based on the parsed R source code, generate corresponding bytecode at least of a first Java class and transfer same to the Java runtime environment.

### Claim 2 (depends on 1)

The device according to claim 1, further configured to store the bytecode of the at least one first generated Java class in a library with bytecode of compiled Java classes and, upon a subsequent request for the predetermined R source code, transfer the compiled bytecode, stored in the library, of the first Java class to the Java runtime environment.

### Claim 3 (depends on 1)

The device according to claim 1, further configured to compile at least some functions of an R system library of an R runtime environment in corresponding bytecode of Java classes or retrieve corresponding bytecode of already-compiled Java classes from a library.

### Claim 4 (depends on 1)

The device according to claim 1, which provides at least a second and a third Java class, wherein instances of the second Java class represent a current state of an R runtime environment, instances of the third Java class represent objects of the R runtime environment and a plurality of instances of the third Java class can be allocated to one instance of the second Java class.

### Claim 5 (depends on 4)

The device according to claim 4, wherein the at least one first Java class comprises at least a method with a transfer parameter with a type in accordance with the second Java class and/or a method with a result type in accordance with the second Java class so that information of an R runtime environment can be transferred as instances of the third Java class to/from the at least one method.

### Claim 6 (depends on 1)

A cluster system having a plurality of mutually coupled cluster nodes that process extensive data, wherein each of the cluster nodes has a Java runtime environment, at least one Java runtime environment has a class loading device according to claim 1 and is configured to compile R source code of at least one function using the class loading device in corresponding bytecode of a first Java class.

### Claim 7 (depends on 6)

The cluster system according to claim 6, wherein each cluster node further comprises a Java framework to implement a MapReduce algorithm and access a distributed file system to store data on the plurality of cluster nodes.

### Claim 8 (depends on 7)

The cluster system according to claim 7, wherein the Java runtime environment is configured to inject a Java method of accessing the distributed file system into the corresponding executable first Java class during execution of the at least one function of the R runtime environment.

### Claim 9

A method of executing a function written in R programming language comprising:
at least one Java class loading device compiling R source code, written in the R programming language, of the function in bytecode at least of a first Java class with a method corresponding to the function; and
executing the function within the Java runtime environment by calling the bytecode of the corresponding method of the first Java class.

### Claim 10 (depends on 9)

The method according to claim 9, wherein the corresponding method of the first Java class expects at least one Java object in accordance with a second Java class as a parameter, the at least one Java object represents a current state of an R runtime environment, and the corresponding method of the first Java class is executed using an instance of the second Java class as a parameter.

### Claim 11 (depends on 7)

The cluster system according to claim 7, wherein the Java framework is a Hadoop framework.

### Claim 12 (depends on 7)

The cluster system according to claim 7, wherein the cluster nodes are a Hadoop Distributed File System (HDFS).
