Robert Hooke provides a flexible, composable mechanism by which you can extend behaviour of functions after they've been defined. It's named after Robert Hooke FRS, a founding member of the Royal Society who made many important discoveries in the fields of Gravitation, Microscopy, and Astronomy.
Add this to your project.clj :dependencies
list:
[robert/hooke "1.3.0"]
If you would like to make your software extensible using Hooke, all you need to do is provide a convention for namespaces that will get loaded on startup. Then users can place files that call add-hook under a specific namespace prefix (my.program.hooks.*) which they can rely on getting loaded at startup.
Hooks can change the behaviour of the functions they wrap in many ways:
Hooke is inspired by Emacs Lisp's defadvice and clojure.test fixtures.
(use 'robert.hooke)
(defn examine [x]
(println x))
(defn microscope
"The keen powers of observation enabled by Robert Hooke allow
for a closer look at any object!"
[f x]
(f (.toUpperCase x)))
(defn doubler [f & args]
(apply f args)
(apply f args))
(defn telescope [f x]
(f (apply str (interpose " " x))))
(add-hook #'examine #'microscope)
(add-hook #'examine #'doubler)
(add-hook #'examine #'telescope)
(examine "something")
> S O M E T H I N G
> S O M E T H I N G
Hooks are functions that wrap other functions. They receive the original function and its arguments as their arguments. Hook functions can wrap the target functions in binding, change the argument list, only run the target functions conditionally, or all sorts of other stuff.
Technically the first argument to a hook function is not always the target function; if there is more than one hook then the first hook will receive a function that is a composition of the remaining hooks. (Dare I say a continuation?) But when you're writing hooks, you should act as if it is the target function.
Adding hooks to a defmulti is discouraged as it will make it impossible to add further methods. Hooks are meant to extend functions you don't control; if you own the target function there are obviously better ways to change its behaviour.
When adding hooks it's best to use vars instead of raw functions in order to allow the code to be reloaded interactively. If you recompile a function, it will be re-added as a hook, but if you use a var it will be able to detect that it's the same thing across reloads and avoid duplication.
(add-hook #'some.ns/target-var #'hook-function)
instead of:
(add-hook #'some.ns/target-var hook-function)
Most of the time you'll never need more than just add-hook. But there's more!
If you are using Hooke just to add side-effects to a function, it may
be simpler to use the append
or prepend
macros:
(prepend print-name
(print "The following person is awesome:"))
(print-name "Gilbert K. Chesterton")
> The following person is awesome:
> Gilbert K. Chesterton
You may also run a block of code with the hooks for a given var stripped out:
(with-hooks-disabled print-name
(print-name "Alan Moore"))
> Alan Moore
The with-scope
macro provides a scope which records any change to hooks during
the dynamic scope of its body, and restores hooks to their original state on
exit of the scope. Note that all threads share the scope. Using the example
functions above:
(examine "something")
> something
(with-scope
(add-hook #'examine #'microscope)
(examine "something"))
> SOMETHING
(examine "something")
> something
Copyright © 2010-2012 Phil Hagelberg, Kevin Downey, and contributors.
Distributed under the Eclipse Public License, the same as Clojure.
Can you improve this documentation? These fine people already did:
Phil Hagelberg, Hugo Duncan, Paul Dorman & Joe GalloEdit on GitHub
cljdoc is a website building & hosting documentation for Clojure/Script libraries
× close