On this page
Creating a subprocess
Concepts Jump to heading
- Deno is capable of spawning a subprocess via Deno.Command.
--allow-runpermission is required to spawn a subprocess.- Spawned subprocesses do not run in a security sandbox.
- Communicate with the subprocess via the stdin, stdout and stderr streams.
Simple example Jump to heading
This example is the equivalent of running echo "Hello from Deno!" from the
command line.
// define command used to create the subprocess
const command = new Deno.Command("echo", {
args: [
"Hello from Deno!",
],
});
// create subprocess and collect output
const { code, stdout, stderr } = await command.output();
console.assert(code === 0);
console.log(new TextDecoder().decode(stdout));
console.log(new TextDecoder().decode(stderr));
Run it:
$ deno run --allow-run=echo ./subprocess_simple.ts
Hello from Deno!
Security Jump to heading
The --allow-run permission is required for creation of a subprocess. Be aware
that subprocesses are not run in a Deno sandbox and therefore have the same
permissions as if you were to run the command from the command line yourself.
Communicating with subprocesses Jump to heading
By default when you use Deno.Command() the subprocess inherits stdin,
stdout and stderr of the parent process. If you want to communicate with a
started subprocess you must use the "piped" option.
Piping to files Jump to heading
This example is the equivalent of running yes &> ./process_output in bash.
import {
mergeReadableStreams,
} from "jsr:@std/streams@1.0.0-rc.4/merge-readable-streams";
// create the file to attach the process to
const file = await Deno.open("./process_output.txt", {
read: true,
write: true,
create: true,
});
// start the process
const command = new Deno.Command("yes", {
stdout: "piped",
stderr: "piped",
});
const process = command.spawn();
// example of combining stdout and stderr while sending to a file
const joined = mergeReadableStreams(
process.stdout,
process.stderr,
);
// returns a promise that resolves when the process is killed/closed
joined.pipeTo(file.writable).then(() => console.log("pipe join done"));
// manually stop process "yes" will never end on its own
setTimeout(() => {
process.kill();
}, 100);
Run it:
$ deno run --allow-run=yes --allow-read=. --allow-write=. ./subprocess_piping_to_file.ts
Reading subprocess output with convenience methods Jump to heading
When working with spawned subprocesses, you can use convenience methods on the
stdout and stderr streams to easily collect and parse output. These methods
are similar to those available on Response objects:
const command = new Deno.Command("deno", {
args: [
"eval",
"console.log(JSON.stringify({message: 'Hello from subprocess'}))",
],
stdout: "piped",
stderr: "piped",
});
const process = command.spawn();
// Use convenience methods to collect output
const stdoutText = await process.stdout.text();
const stderrText = await process.stderr.text();
console.log("stdout:", stdoutText);
console.log("stderr:", stderrText);
// Wait for the process to complete
const status = await process.status;
console.log("Exit code:", status.code);
Available convenience methods include:
.text()- Returns the output as a UTF-8 string.bytes()- Returns the output as aUint8Array.arrayBuffer()- Returns the output as anArrayBuffer.json()- Parses the output as JSON and returns the parsed object
const command = new Deno.Command("deno", {
args: ["eval", "console.log(JSON.stringify({name: 'Deno', version: '2.0'}))"],
stdout: "piped",
});
const process = command.spawn();
// Parse JSON output directly
const jsonOutput = await process.stdout.json();
console.log("Parsed JSON:", jsonOutput); // { name: "Deno", version: "2.0" }
await process.status;