process

A Clojure wrapper around java.lang.ProcessBuilder.

Status: alpha.

This library is included in babashka since 0.2.3 but is also intended as a JVM library:

$ clojure -Sdeps '{:deps {babashka/process {:mvn/version "0.0.1"}}}'

user=> (require '[clojure.string :as str])
nil
user=> (require '[babashka.process :refer [$ check]])
nil
user=> (-> ^{:out :string} ($ ls -la) check :out str/split-lines first)
"total 136776"

Differences with clojure.java.shell/sh

If clojure.java.shell works for your purposes, keep using it. But there are contexts in which you need more flexibility. The major differences compared with this library:

• sh is blocking, process makes blocking explicit via deref
• sh focuses on convenience but limits what you can do with the underlying process, process exposes as much as possible while still offering an ergonomic API
• process supports piping processes via -> or pipeline
• sh offers integration with clojure.java.io/copy for :in, process extends this to :out and :err

API

You will probably mostly need process or $ and check so it would be good to start reading the docs for these. Skim over the rest and come back when you need it.

• process: takes a command (vector of strings or objects that will be turned into strings) and optionally a map of options.

  Returns: a record (called "the process" in this README) with

  • :proc: an instance of java.lang.Process
  • :in, :err, :out: the process's streams. To obtain a string from :out or :err you will typically use slurp or use the :string option (see below). Slurping those streams will block the current thread until the process is finished.
  • :cmd: the command that was passed to create the process.
  • :prev: previous process record in case of a pipeline.

  The returned record can be passed to deref. Doing so will cause the current thread to block until the process is finished and will populate :exit with the exit code.

  Supported options:

  • :in, :out, :err: objects compatible with clojure.java.io/copy that will be copied to or from the process's corresponding stream. May be set to :inherit for redirecting to the parent process's corresponding stream.Optional :in-enc, :out-enc and :err-enc values will be passed along to clojure.java.io/copy.

    The :out and :err options support :string for writing to a string output. You will need to deref the process before accessing the string via the process's :out.

  • :inherit: if true, sets :in, :out and :err to :inherit.

  • :dir: working directory.

  • :env, :extra-env: a map of environment variables. See Add environment.

  • :escape: function that will applied to each stringified argument. On Windows this defaults to prepending a backslash before a double quote. On other operating systems it defaults to identity.

  • :shutdown: shutdown hook, defaults to nil. Takes process map. Typically used with destroy or destroy-tree to ensure long running processes are cleaned up on shutdown.

  Piping can be achieved with the -> macro:

  (-> (process '[echo hello]) (process '[cat]) :out slurp) ;;=> "hello\n"
  

  or using the pipeline function (see below)

• check: takes a process, waits until is finished and throws if exit code is non-zero.

• $: convenience macro around process. Takes command as varargs. Options can be passed via metadata on the form. Supports interpolation via ~.

• sh: convenience function similar to clojure.java.shell/sh that sets :out and :err to :string by default and blocks. Similar to cjs/sh it does not check the exit code (this can be done with check).

• *defaults*: dynamic var containing overridable default options. Use alter-var-root to change permanently or binding to change temporarily.

• destroy: function of process or map with :proc (java.lang.ProcessBuilder). Destroys the process and returns the input arg.

• destroy-tree: same as destroy but also destroys all descendants. JDK9+ only.

• pb: returns a process builder (as record).

• start: takes a process builder, calls start and returns a process (as record).

• pipeline:

  • When passing a process, returns a vector of processes of a pipeline created with -> or pipeline.
  • When passing two or more process builders created with pb: creates a pipeline as a vector of processes (JDK9+ only).

  Also see Pipelines.

Usage

user=> (require '[babashka.process :refer [process $ check]])

Invoke ls:

user=> (-> (process '[ls]) :out slurp)
"LICENSE\nREADME.md\nsrc\n"

Change working directory:

user=> (-> (process '[ls] {:dir "test/babashka"}) :out slurp)
"process_test.clj\n"

Set the process environment.

user=> (-> (process '[sh -c "echo $FOO"] {:env {:FOO "BAR" }}) :out slurp)
"BAR\n"

The return value of process implements clojure.lang.IDeref. When dereferenced, it will wait for the process to finish and will add the :exit value:

user=> (-> @(process '[ls foo]) :exit)
Execution error (ExceptionInfo) at babashka.process/check (process.clj:74).
ls: foo: No such file or directory

The function check takes a process, waits for it to finish and returns it. When the exit code is non-zero, it will throw.

user=> (-> (process '[ls foo]) check :out slurp)
Execution error (ExceptionInfo) at babashka.process/check (process.clj:74).
ls: foo: No such file or directory

Redirect output to stdout:

user=> (do (process '[ls] {:out :inherit}) nil)
LICENSE		README.md	deps.edn	src		test
nil

Both :in, :out may contain objects that are compatible with clojure.java.io/copy:

user=> (with-out-str (check (process '[cat] {:in "foo" :out *out*})))
"foo"

user=> (with-out-str (check (process '[ls] {:out *out*})))
"LICENSE\nREADME.md\ndeps.edn\nsrc\ntest\n"

The :out option also supports :string. You will need to deref the process in order for the string to be there:

user=> (-> @(process '[ls] {:out :string}) :out)
"LICENSE\nREADME.md\ndeps.edn\nsrc\ntest\n"

Redirect output stream from one process to input stream of the next process:

(let [is (-> (process '[ls]) :out)]
  (process ["cat"] {:in is
                    :out :inherit})
    nil)
LICENSE
README.md
deps.edn
src
test
nil

Forwarding the output of a process as the input of another process can also be done with thread-first:

(-> (process '[ls])
    (process '[grep "README"]) :out slurp)
"README.md\n"

Demo of a cat process to which we send input while the process is running, then close stdin and read the output of cat afterwards:

(ns cat-demo
  (:require [babashka.process :refer [process]]
            [clojure.java.io :as io]))

(def catp (process '[cat]))

(.isAlive (:proc catp)) ;; true

(def stdin (io/writer (:in catp)))

(binding [*out* stdin]
  (println "hello"))

(.close stdin)

(slurp (:out catp)) ;; "hello\n"

(def exit (:exit @catp)) ;; 0

(.isAlive (:proc catp)) ;; false

$ and sh

$ is a convenience macro around process:

(def config {:output {:format :edn}})
(-> ($ clj-kondo --config ~config --lint "src") deref :out slurp edn/read-string)
{:findings [], :summary {:error 0, :warning 0, :info 0, :type :summary, :duration 34}}

sh is a convenience function around process which sets :out and :err to :string and blocks automatically, similar to clojure.java.shell/sh and unlike $:

(def config {:output {:format :edn}})
(-> (sh ["clj-kondo" "--config" config" "--lint" "src"]) :out slurp edn/read-string)
{:findings [], :summary {:error 0, :warning 0, :info 0, :type :summary, :duration 34}}

Tokenization

Both process, $ and sh support tokenization when passed a single string argument:

(-> (sh "echo hello there") :out)
"hello there\n"

(-> (sh "clj-kondo --lint -" {:in "(inc)"}) :out print)
<stdin>:1:1: error: clojure.core/inc is called with 0 args but expects 1
linting took 11ms, errors: 1, warnings: 0

Output buffering

Note that check will wait for the process to end in order to check the exit code. When the process has lots of data to write to stdout, it is recommended to add an explicit :out option to prevent deadlock due to buffering. This example will deadlock because the process is buffering the output stream but it's not being consumed, so the process won't be able to finish:

user=> (-> (process ["cat"] {:in (slurp "https://datahub.io/datahq/1mb-test/r/1mb-test.csv")}) check :out slurp count)

The way to deal with this is providing an explicit :out option so the process can finish writing its output:

user=> (-> (process ["cat"] {:in (slurp "https://datahub.io/datahq/1mb-test/r/1mb-test.csv") :out :string}) check :out count)
1043005

Add Environment

The :env option replaces your entire environment with the provided map. To add environment variables you can use :extra-env instead:

:extra-env {"FOO" "BAR"}

Pipelines

The pipeline function returns a sequential of processes from a process that was created with -> or by passing multiple objects created with pb:

(mapv :cmd (pipeline (-> (process '[ls]) (process '[cat]))))
[["ls"] ["cat"]]

(mapv :cmd (pipeline (pb '[ls]) (pb '[cat])))
[["ls"] ["cat"]]

To obtain the right-most process from the pipeline, use last (or peek):

(-> (pipeline (pb ["ls"]) (pb ["cat"])) last :out slurp)
"LICENSE\nREADME.md\ndeps.edn\nsrc\ntest\n"

Calling pipeline on the right-most process returns the pipeline:

(def p (pipeline (pb ["ls"]) (pb ["cat"])))
#'user/p
(= p (pipeline (last p)))
true

To check an entire pipeline for non-zero exit codes, you can use:

(run! check (pipeline (-> (process '[ls "foo"]) (process '[cat]))))
Execution error (ExceptionInfo) at babashka.process/check (process.clj:37).
ls: foo: No such file or directory

Although you can create pipelines with ->, for some applications it may be preferable to create a pipeline with pipeline which defers to ProcessBuilder/startPipeline. In the following case it takes a long time before you would see any output due to buffering.

(future
  (loop []
    (spit "log.txt" (str (rand-int 10) "\n") :append true)
    (Thread/sleep 10)
    (recur)))

(-> (process '[tail -f "log.txt"])
    (process '[cat])
    (process '[grep "5"] {:out :inherit}))

The solution then it to use pipeline + pb:

(pipeline (pb '[tail -f "log.txt"])
          (pb '[cat])
          (pb '[grep "5"] {:out :inherit}))

The varargs arity of pipeline is only available in JDK9 or higher due to the availability of ProcessBuilder/startPipeline. If you are on JDK8 or lower, the following solution that reads the output of tail line by line may work for you:

(def tail (process '[tail -f "log.txt"] {:err :inherit}))

(def cat-and-grep
  (-> (process '[cat]      {:err :inherit})
      (process '[grep "5"] {:out :inherit
                            :err :inherit})))

(binding [*in*  (io/reader (:out tail))
          *out* (io/writer (:in cat-and-grep))]
  (loop []
    (when-let [x (read-line)]
      (println x)
      (recur))))

Another solution is to let bash handle the pipes by shelling out with bash -c.

Notes

Script termination

Because process spawns threads for non-blocking I/O, you might have to run (shutdown-agents) at the end of your Clojure JVM scripts to force termination. Babashka does this automatically.

Clojure.pprint

When pretty-printing a process, you will get an exception:

(require '[clojure.pprint :as pprint])
(pprint/pprint (process ["ls"]))
Execution error (IllegalArgumentException) at user/eval257 (REPL:1).
Multiple methods in multimethod 'simple-dispatch' match dispatch value: class babashka.process.Process -> interface clojure.lang.IDeref and interface clojure.lang.IPersistentMap, and neither is preferred

The reason is that a process is both a record and a clojure.lang.IDeref and pprint does not have a preference for how to print this. This can be resolved using:

(prefer-method pprint/simple-dispatch clojure.lang.IPersistentMap clojure.lang.IDeref)

License

Copyright © 2020 Michiel Borkent

Distributed under the EPL License. See LICENSE.