Execute rJava code in a separate process with specific Java version

Runs the provided function in a fresh R subprocess where rJava has not yet been loaded. This allows you to enforce a specific Java version for rJava operations without affecting the main R session or requiring a restart.

Usage

with_rjava_env(
  version,
  func,
  args = list(),
  distribution = "Corretto",
  install = TRUE,
  accept_system_java = TRUE,
  quiet = TRUE,
  libpath = .libPaths()
)

Arguments

version: Integer or character. Required. The Java version you need (e.g., 17, 21). Defaults to NULL, which is invalid and will trigger a validation error; callers should always provide a non-NULL value explicitly.
func: The function to execute in the subprocess.
args: A list of arguments to pass to func.
distribution: Character. The Java distribution to download. Defaults to "Corretto".
install: Logical. If TRUE (default), attempts to download/install if missing. If FALSE, returns FALSE if the version is not found.
accept_system_java: Logical. If TRUE (default), the function will scan the system for existing Java installations (using JAVA_HOME, PATH, and OS-specific locations). If a system Java matching the version and type requirements is found, it will be used. Set to FALSE to ignore system installations and strictly use an rJavaEnv managed version.
quiet: A logical value indicating whether to suppress messages. Can be TRUE or FALSE.
libpath: Optional character vector of library paths to use in the subprocess. Defaults to .libPaths().

Value

The result of func.

Details

This function requires the callr package.

Examples

if (FALSE) { # \dontrun{
# Run a function using Java 21 in a subprocess
result <- with_rjava_env(
  version = 21,
  func = function(x) {
    library(rJava)
    .jinit()
    .jcall("java.lang.System", "S", "getProperty", "java.version")
  },
  args = list(x = 1)
)
print(result)
} # }