![What is Clojure?
Hosted
Functional
First class functions
Build software by composing over functions
EXAMPLE 2
Applying arbitrary function to a collection of data
!
;;
Name
a
collection
of
numbers
(def
data
[1
2
3
4
5])
!
;;
Creates
a
function
called
process-‐data
that
accepts
a
collection
and
an
operation
to
be
(defn
process-‐data
[collection
operation]
(apply
operation
collection))
!
;;
Addition
(process-‐data
data
+)
!
;;
Multiplication
(process-‐data
data
*)
!
Relax if you aren’t following the syntax :-)
You’ll learn this in the workshop](https://image.slidesharecdn.com/atasteofclojure-150411141225-conversion-gate01/85/A-Taste-of-Clojure-4-320.jpg)
![What is Clojure?
Hosted
Functional
Immutability
Easy Reasoning and Debugging
In the following Ruby code, can you be sure
that x is not modified?
!
x
=
{problems:
99}
!
dodgyMethod(x);
!
#
what
is
the
value
of
x?
x[":problems"]](https://image.slidesharecdn.com/atasteofclojure-150411141225-conversion-gate01/85/A-Taste-of-Clojure-5-320.jpg)
![What is Clojure?
Hosted
Functional
Parallelism
Concurrency
Lisp
Collection Types
Collection types in Clojure
!
List (a b c d e) Essentially a linked list
Vector [a b c d e] Like “arrays” in Ruby/Python
!
Wait a second…
Didn’t we already see those square brackets and
parentheses when we define a function?](https://image.slidesharecdn.com/atasteofclojure-150411141225-conversion-gate01/85/A-Taste-of-Clojure-20-320.jpg)
![What is Clojure?
Hosted
Functional
Parallelism
Concurrency
Lisp
Homoiconicity
Homoiconicity
Lisp is a language where the structure of the code is
represented by standard data structures (list, vectors,
hash).
Repeat after me
!
Code as Data. Data as Code.
!
!
(defn
wont-‐make-‐you-‐blind
[name]
(println
name
",
parentheses
won't
make
you
blind"))](https://image.slidesharecdn.com/atasteofclojure-150411141225-conversion-gate01/85/A-Taste-of-Clojure-21-320.jpg)
A Taste of Clojure
- 1. A taste of ! ! David Leung! ! Twitter: @davleung Email: david@davidslab.com GPG Fingerprint: 217E 1ECE 2349 D178 73E5 C194 1E2E C02A A74A A531
- 2. What is Clojure? Hosted Runs on multiple major platforms JVM CLR JS Runtime (ClojureScript)
- 3. What is Clojure? Hosted Excellent Interops EXAMPLE 1 Retrieve a webpage using Java library ! ! ;; Imports the java.net.URL class into current namespace (import ‘java.net.URL) ;; Creating an instance of URL. ;; In Java: ;; URL con = new URL(“http://www.reddit.com”) (def con (URL. “http://www.reddit.com”)) ! ! ;; GET the page and print it! (println (slurp con))
- 4. What is Clojure? Hosted Functional First class functions Build software by composing over functions EXAMPLE 2 Applying arbitrary function to a collection of data ! ;; Name a collection of numbers (def data [1 2 3 4 5]) ! ;; Creates a function called process-‐data that accepts a collection and an operation to be (defn process-‐data [collection operation] (apply operation collection)) ! ;; Addition (process-‐data data +) ! ;; Multiplication (process-‐data data *) ! Relax if you aren’t following the syntax :-) You’ll learn this in the workshop
- 5. What is Clojure? Hosted Functional Immutability Easy Reasoning and Debugging In the following Ruby code, can you be sure that x is not modified? ! x = {problems: 99} ! dodgyMethod(x); ! # what is the value of x? x[":problems"]
- 6. What is Clojure? Hosted Functional Immutability Separation of Identity & State ! Identity is a reference to something. State is the aggregated values of an identity at a particular point in time. In Clojure, an identity’s state is updated by creating a new value and assign it back to the identify. ! Whaaaaat? Isn’t that inefficient???
- 7. What is Clojure? Hosted Functional Immutability Efficient Structure Sharing ! Let’s create a hash: !(def a {:a 5 :b 6 :c 7 :d 8}) ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! Credit: Image from “Clojure Programming” by Chas Emerick, Brian Carper & Christophe Grand
- 8. What is Clojure? Hosted Functional Immutability Efficient Structure Sharing ! !(def a {:a 5 :b 6 :c 7 :d 8}) ! (def b (assoc a :c 0)) ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! “Copying” data is cheap! ! ! ! Credit: Image from “Clojure Programming” by Chas Emerick, Brian Carper & Christophe Grand
- 9. What is Clojure? Hosted Functional Immutability Fosters Concurrent Programming ! “Immutable objects are always thread safe.” —Brian Goetz (Author of Java Concurrency in Practice) ! ! ! More about Concurrency later
- 10. What is Clojure? Hosted Functional Parallelism Parallel processing Trivial Parallel Processing Problem Problem The Boss just asked you to parse 100K serialized JSON objects into a hash. Each string is ~300KB. The collection/array of strings are stored in the variable problemsCollection. ! Write a snippet that makes use of all cores to process this collection. Don’t worry about storing the result. ! Do this within a minute, or you are fired. ! Time starts now.
- 11. What is Clojure? Hosted Functional Parallelism Parallel processing Trivial Parallel Processing Problem Solution ! ! ;; non-‐parallelized version that will get you fired (dorun (map json/read-‐str problemsCollection)) ! ;; parallelized version (dorun (pmap json/read-‐str problemsCollection)) ! ! EXAMPLE 3 (DEMO) ! Note pmap is a parallelized version of map
- 12. What is Clojure? Hosted Functional Parallelism Parallel processing Watch out for the overhead involved in distributing work to the work! ! EXAMPLE 4 (DEMO) !
- 13. What is Clojure? Hosted Functional Parallelism Concurrency Two types of Concurrent Operations Coordinated Concurrency One or more actors (threads) must coordinate to produce the right computation. e.g. Bank transactions ! Synchronous Concurrency Should the thread that wants to update a reference be blocked until their return?
- 14. What is Clojure? Hosted Functional Parallelism Concurrency 3 flavors of Reference Types Coordinated Uncoordinated Synchronous Refs Atoms Asynchronous Agents
- 15. What is Clojure? Hosted Functional Parallelism Concurrency Atoms Create an atom (def counter (atom 0)) Access an atom’s value (dereferencing) (deref counter) ! ;; syntactical sugar @counter Compare and Swap (swap! counter + 2) ! Note: Clojure uses ! in a function’s name to denote that the function mutates state (memory/IO).
- 16. What is Clojure? Hosted Functional Parallelism Concurrency Software Transactional Memory (STM) STM: Database-like Transactions for Concurrency ! Clojure controls updates to references via a system called STM. It removes the need for manual locking. ! ! ! ! ! Fulfills the Atomicity, Consistency, and Isolation aspects of the Database Transaction Properties ACID. STM is to concurrency what Garbage Collector is to memory management
- 17. What is Clojure? Hosted Functional Parallelism Concurrency Software Transactional Memory (STM) There are more to concurrency Clojure has “data flow variables” too! We’ll cover concurrency in the workshops later. Things we will look at: ! Future Delay Promise
- 18. What is Clojure? Hosted Functional Parallelism Concurrency Lisp What is a Lisp? ! ! ! ! ! ! Lisp is invented by John McCarthy in 1958. Lisp pioneered ideas like tree data structures, dynamic typing, garbage collection, higher- order functions, recursion. John McCarthy is quite the hero in Computer Science.
- 19. What is Clojure? Hosted Functional Parallelism Concurrency Lisp Lisp: LISt Processing Lisp is written in symbolic expressions, or S- expressions. S-expressions are written as a pair of parenthesis surrounding a sequence of symbols: (f a b c) ! The first position in the list is also known as the functional position. When evaluated, the function f will be called with the rest of the list as arguments.
- 20. What is Clojure? Hosted Functional Parallelism Concurrency Lisp Collection Types Collection types in Clojure ! List (a b c d e) Essentially a linked list Vector [a b c d e] Like “arrays” in Ruby/Python ! Wait a second… Didn’t we already see those square brackets and parentheses when we define a function?
- 21. What is Clojure? Hosted Functional Parallelism Concurrency Lisp Homoiconicity Homoiconicity Lisp is a language where the structure of the code is represented by standard data structures (list, vectors, hash). Repeat after me ! Code as Data. Data as Code. ! ! (defn wont-‐make-‐you-‐blind [name] (println name ", parentheses won't make you blind"))
- 22. What is Clojure? Hosted Functional Parallelism Concurrency Lisp Homoiconicity Construct a function call with collection manipulation functions ! ;; the quotation mark stops Clojure treating the ;; list as a function call (def arguments '(1 2 3)) ! ;; cons is a traditional lisp function ;; that prepends an object to the head of a list (def function-‐call (cons + arguments)) ! (eval function-‐call) ;; => 6 ! !
- 23. What is Clojure? Hosted Functional Parallelism Concurrency Lisp Macros Building Malleable Systems ! As well as being able to write code that “writes code” (metaprogramming), you can even change the evaluation semantics by writing macros! We’ll explore this in the workshops.
- 24. What is Clojure? Hosted Functional Parallelism Concurrency Lisp Language Extensions The flexible of Clojure allows Clojure and your codebase to stretch and bend to your needs. Notable Clojure extensions: ! core.async Implementation of channels and blocks from the Go programming language. core.logic Implementation of miniKanren for doing relational/constraint logic programming. core.typed Optional typing for Clojure.
- 25. Working with Clojure Editors ! emacs Light Table VimClojure CounterClockWise (eclipse) IntelliJ
- 26. Working with Clojure Build Tools ! Leiningen for automating Clojure projects without setting your hair on fire http://leiningen.org ! ht
- 27. Working with Clojure Learning Resources Test Driven Learning 4Clojure 4clojure.com Clojure Koans clojurekoans.com ! Books Clojure Programming Joy of Clojure Second Edition Clojure Cookbook
- 28. Working with Clojure ! ! David Leung! ! Twitter: @davleung Email: david@davidslab.com GPG Fingerprint: 217E 1ECE 2349 D178 73E5 C194 1E2E C02A A74A A531