Clojure

(Now former) Java developer sneaks up on Clojure

View project on GitHub

Written in January 2014, with Clojure 1.5.1 and Leiningen 2.3.4 on Java 1.7.0_45.
Inspired by Aphyr's excellent, Clojure from-the-ground-up tutorial and made possible by "The joy of Clojure" book. I also heavily used great "Clojure for Java programmers" talks by Rich Hickey, father of Clojure.

Inspired? Made possible? Who am I kidding, I flat out stole from those kind people.

Clojure? That's a Lisp, for god's sake!

I've been avoiding Clojure for a long time. Because it's a Lisp dialect and I have a negative Lisp experience from school. It was either a bad timing or a bad teacher, I guess.
Hmm, I guess my guessing about it, in itself, implies the answer :)

Times and times again it gets apparent to me. Late puberty and adolescence are not the right times to go wide with programming languages!

:)

It must had something to do with the way the book was written

My renewed interest in Clojure was due to a chance encounter. I stumbled upon Aphyr's (Kyle Kingsbury) fascinating Jepsen series, a blog about perils of uncertainty in distributed systems.

There, Clojure looked terse and concise, yet expressive and simple. So I decided to give it a shot. I picked up The joy of Clojure, got in my sweatshirt, put on a headband, closed my wife and our two-month-old twins in the living room, slightly licked my finger and opened up the first page...

After the initial discovery phase, as I was going through the book, I suddenly found myself infatuated with the language. It's beautifully uniform, consistent and simple (which is different from "easy", mind you). At that point, it has just spread out to me, ready to be gulped away. And that's a great feeling!

How would you describe Clojure?

Clojure prides itself in being a dynamic functional programming language.
It is built on three great facilities, immutable data, first-class functions and dynamic typing.

Immutable data means that a function always produces the same output, given the same input and that functions are side-effects free (when a function is run, nothing changes outside that function). All collections are immutable by design. Immutability in Clojure is not optional, like for instance, in Scala.

First class functions means that a variable can be bound to a function, that a function can be passed to another (higher-order) function or that it may be returned from a function. In short, function is data.

Dynamic typing means that we don't declare types. Data types are resolved at runtime.

So from now on, when we say "modify", "add", "remove", ... it really means - create a "copy", modify the copy and return a reference to that new object. But the word "copy" from the previous sentence doesn't stand for a full, brute-force copy. In order to provide immutability and still preserve the performance guarantees (big O notation), Clojure uses something called Structural sharing, which basically means that the data structure of the new object is built by creating references to the elements of the old object, varying only with respect to the modified elements (called path copying).

To picture this, think of a tree that gets a new node. Nodes of the new tree simply point to nodes of the old tree, so that, consequently, only a single full-blown node object gets created.

Set the fuck up!

And gimme PRs

Leiningen logo

Disclaimer: I write this as I'm going through the book myself, so bear with me. Open pull requests as you see fit

OK, that's more than enough BS (for now). Let's start by setting up our Clojure runtime environment.

To set it up, I suggest you use an excellent automation tool (dependency mgr, builder, test runner, packager, all-in-one) Leiningen.

Notice there's no "installer" in the list of features. That's because Clojure is just another dependency of "our project".

For your convenience, here's a dead simple, step-by-step environment setup for Ubuntu (other OS, sir?):


cd /usr/bin
sudo curl -O https://raw.githubusercontent.com/technomancy/leiningen/stable/bin/lein
sudo chmod a+x lein

Let me explain what we just did. First we fetched lein script with curl° into /usr/bin°°. We made sure it's executable with chmod.

° Use sudo apt-get install curl, if you haven't installed it yet (doesn't come OOTB with Ubuntu).
°° You can use any other dir, as long as you make sure it's in your PATH (check with echo $PATH), which makes possible to run it from anywhere.

Let's kick things off:

cd
lein new scratch

Here we used lein new to create a fresh Clojure project, based on the scratch template (scratch - as in scratch the surface, I guess). Lein, in turn, installed rest of itself into ~/.lein/self-installs.

In case you get stuck, visit lein install instructions.
If you want to know more, the official Leiningen tutorial can be found right here.

Lovely! Now fire along:

cd scratch
lein repl

Boom! We have our own working Clojure environment.

repl

REPL is short for Read Eval Print Loop, which is like an interactive window into a Clojure program (similar to a JavaScript console in the browser or REPL in Node).

Now when that's all sorted and you're eager to learn, let's see how Clojure code looks like...

Basics

A valid Clojure expression consists of numbers, symbols, lists, keywords, booleans, characters, functions, function calls, macros (wtf?), strings, literal maps, vectors, and sets.

All of those, except symbols and lists, evaluate to themselves.

Symbols are similar to variables in other languages, which basically means that, when a symbol is encountered, compiler tries to find the value that the symbol was previously (hopefully) bound to, which is then used in place of the symbol.

Lists take a special place in Clojure (after all, Lisp means LISt Processing). They are the main building block of the language. Lists start with a so called operation form:

(op ...)

op can be either:

  • one of very few special operations (listed bellow)
  • a macro
  • an expression that yields a function

For instance, if the Clojure compiler encounters:

(my-fun some-expr)

it'll first try to resolve my-fun, which in our case, is a previously defined function that conceptually looks like this:

(def my-fun value-expr)

Then, it'll evaluate some-expr and invoke the my-fun function, passing in the result.

so my-fun is a symbol that is bound to a function definition

Full list of special operations:

def     ;evaluates an expression and binds the result to a symbol
if      ;conditional evaluation
fn      ;defines a function
let     ;establishes a name in a local lexical scope
loop    ;used for functional looping
recur   ;used to support recursion in functional looping
do      ;defines a block of statements
new     ;allocates a new java object
.       ;used to access a java method
throw   ;same as in java
try     ;same as in java
set!    ;re-binds a symbol
quote   ;supresses evaluation of an expression (same as single quote, ')
var     ;provides a mechanism to refer to a mutable storage location inside a thread (?)

The list above may contain some descriptions that may not be clear to you, but soon, if you stuck with it, it'll all get cleared. And that's a promise.

Here's how a function invocation looks like:

user=> (inc 2)      ;"increment 2"
3                   ;this is REPL output
;=> 3               ;which I'll write like this from now on

; those three that begin with a semicolon are comments
;;this one also, from that first semicolon to the end of the line. All the way here ->

Syntax

Prefix/Polish notation

;; intentionally skipping java semicolons

;;------------------------------------
;;     Java             Clojure
;;------------------------------------
int i = 5              (def i 5)
;;------------------------------------
if(x == 0)             (if (zero? x)
  return y               y
else                     z)
  return z 
;;------------------------------------
x * y * z              (* x y z)
;;------------------------------------
foo(x, y, z)           (foo x y z)
;;------------------------------------

You can try the following example in your REPL:

;skipping the 'user=>' prompt from now on
(+ 1 (- 5 2) (+ 3 4))
;=> 11

Uh, that looks somewhat weird, right?

This type of notation, inherited from Lisp, called a prefix notation or Polish notation, may look weird at the first glance.

Let's back up a bit.
All programming languages, in order to execute the source code, need to parse it first. In most languages, the product of this code parsing is a so called abstract syntax tree (AST), which is then fed into the compiler.

Let's see how that tree looks for the example at hand.
In Java, the expression above would be written like this:

1 + (5 - 2) + (3 + 4);   // parentheses left for clarity
// 1 + 3 + 7
// 11

... and the AST would look like this:

ast

After seeing what AST (as the ideal structure for representing code) looks like, I argue that prefix is the natural way of representing expressions.

When you think of it (really hard), as early as first grade maths, the only option we ever see is infix notation, so that's what gets hardwired inside our brains.

That is why, IMO, the Polish notation looks weird to us.

prefix notation allows any number of arguments in an operation (infix only two). Moreover, it completely eliminates the problem of operator precedence

at this point I have to stress that this is my opinion after only a couple of weeks learning Clojure. I'm slightly affraid of what would happen to me after I explore all corners of the language :)

Clojure is different

This is how a program is executed in a traditional, java-style edit-compile-run way:

ast

In Java (and it's similar in other traditional languages), the source code gets handed to the compiler, which compiles it to bytecode. The bytecode is then run on the JVM. If you need to change something in your code, e.g. fix a bug, you need to do the whole process all over again. Open up the source file, make some changes, compile the source code, then stop the whole program and finally send the new version of bytecode to the JVM to be executed.

Uh, we got over that hurdle. Now we confidently start our program but we soon notice that it's still not right. OK, we stop the whole program once more and then start it in a debug mode, carefully setting breakpoints along the way. But sometimes it's not that easy, or even possible to replicate the exact context where our bug has previously occurred.

at work, we often used to spend a better half of our day in this iterative process, doing nothing. When you combine that with IBM's tooling, that becomes a nightmare. E.g. Rational IDE weighs more than a GB. Websphere application server takes ages to start. When a new developer needs to set up his environment, it has known to take a couple of days to wire all the stuff together. We are crazy! How on earth we got eased into this unproductive way of working.

ast

Clojure's evaluation is much more dynamic, the code first goes to Reader, which takes the characters and turns them into Clojure data structures. The Compiler never sees the source code, it compiles the data structures produced by the Reader to bytecode, which is then run on the JVM.

So what does this mean? Can such a seemingly subtle difference have a noticeable impact on our development process?

The point is that, this way, we can hot-swap a part of the running code without stopping the program. The source code doesn't have to come from a file. We can use REPL to connect to a running program and evaluate expressions on the fly.

The other benefit is that other programs can easily produce data structures, thus avoiding source code and Reader all together. That's the Clojure's secret sauce. It's what makes macros possible.
When Evaluator, while processing source code, encounters a macro (i.e. a symbol that is bound to a macro), it stops executing that part of the program and sends it to that macro, a little side-program that manipulates data structures, transforming them in some useful way and then returning that, extended data structures back to the Evaluator.

This is extremely powerful concept, which allows us to extend the language without waiting for Rich Hickey to do so. Overwhelming amount of things, that are built-into other languages, are just macros in Clojure. For instance:

(or x y)        ;'or' is a macro in Clojure 

;;after being extended by or macro, becomes:

(let [or__158 x]
  (if or__158 or__158 y))

Don't worry if macros are not clear yet (I explained them rather poorly, I know), we'll dedicate the whole chapter to that powerful construct.

Basics (this time for real)

Scalar data types

A scalar data type is the one that can only hold one value at a time (value of a number, symbol, keyword, string or a character).

Numeric types

The following examples of numeric expressions are trivial, so I suggest you try them out, in your REPL.

(type 3)
;=> java.lang.Long

Long/MAX_VALUE
;=> 9223372036854775807

(inc (bigint Long/MAX_VALUE))
;=> 9223372036854775808N

(type 5N)
;=> clojure.lang.BigInt

(type (int 0))
;=> java.lang.Integer

(type (short 0))
;=> java.lang.Short

(type (byte 0))
;=> java.lang.Byte

;; decimal, hexadecimal, octal, radix-32, and binary literals:
[127 0x7F 0177 32r3V 2r01111111]
;=> [127 127 127 127 127]
;; radix supports up to base 36

(Short/MAX_VALUE)
;=> 32767

(Integer/MAX_VALUE)
;=> 2147483647

Byte/MAX_VALUE
;=> 127

(type 1.23)
;=> java.lang.Double

(type (float 1.23))
;=> java.lang.Float

366e3
;=> 366000.0
(+ 1 2.0)
;=> 3.0

(class (+ 1 2.0))
;=> java.lang.Double

(= 3 3.0)
;=> false

(== 3 3.0)
;=> true

(* 2 3 1/5)
;=> 6/5

(- 5 1 1 1)
;=> 2

(- 2)
;=> -2

(* 4)
;=> 4

(/ 4)
;=> 1/4

(+)
;=> 0     ;neutral for addition

(*)
;=> 1     ;neutral for multiplication

(<= 1 2 3)
;=> true

(<= 1 3 2)
;=> false

(= 2 2 3)
;=> false

(= 2 2 2)
;=> true

To avoid having to round floating-point numbers (e.g. result of 2 / 3), Clojure provides a rational number type clojure.lang.Ratio, thus maintaining absolute precision when dealing with floating point arithmetic:

(type 1/3)
;=> clojure.lang.Ratio

;; rational numbers are automatically simplified, if possible:
100/25
;=> 4

(rational? 2)   ;is rational or may be rational
;=> true

(rational? 2.1)
;=> false

(rationalize 2.1)
;=> 21/10

(numerator (/ 21 10))
;=> 21

(denominator (/ 21 10))
;=> 10

The calculation of rational math, though accurate, isn’t nearly as fast as with floats or doubles (due to overhead cost of e.g. finding the least common denominator).

Truncation

Truncation is a process of limiting the accuracy of floating-point numbers, due to deficiencies in its representation. When a number is truncated, its precision is limited such that the maximum number of digits of accuracy is bound by the number of bits that can fit into the storage space allowed by its representation.

(let [pi-constant 3.14159265358979323846264338327950288419716939937M] ;notice the M
  (println (class pi-constant))
  pi-constant)
;=> java.math.BigDecimal
;=> 3.14159265358979323846264338327950288419716939937M

(let [pi-trunc 3.14159265358979323846264338327950288419716939937]
  (println (class pi-trunc))
  pi-trunc)
;=> java.lang.Double
;=> 3.141592653589793

M, at the end of a floating-point number literal is used to tell Clojure to keep the number in its full precision.
N is used for the same thing when dealing with longs.

Clojure truncates floating point numbers by default

Promotion

Clojure is able to detect when overflow occurs. Then, it automatically promotes the value to a numerical representation that can accommodate larger values

(def small 9)
(class small)
;=> java.lang.Long

(class (+ small 90000000000000000000))
;=> clojure.lang.BigInt

(class (+ small 9.0))
;=> java.lang.Double

(+ Integer/MAX_VALUE Integer/MAX_VALUE)
;=> 4294967294

(class (+ Integer/MAX_VALUE Integer/MAX_VALUE))
;=> java.lang.Long

There's no limit to integer size in Clojure, besides RAM size

Strings and Characters

String is any sequence of characters enclosed in double quotes, including newlines:

"this is a string
on two lines"
;=> "this is a string\non two lines"  ;REPL output includes newline escape

(type "a")
;=> java.lang.String

(str nil)    ;'nil' is Clojure's 'null'
;=> ""

single quote, backtick and quote (though with slightly different properties) are used to include literals in a program without evaluating them

(str 'cat)
;=> "cat"

(str 'a')
;=> "a'"

(str 1)
;=> "1"

(str '1)
;=> "1"

(str true)
;=> "true"

(str '(1 2 3))
;=> "(1 2 3)"

(str "meow " 3 " times")
;=> "meow 3 times"

;characters are denoted by backslash
\a
;=> \a

\u0042
;=> \B

\\
;=> \\

Symbols

Symbols can have either short or full names. The short name is used to refer to things locally, and the fully qualified name is used to refer unambiguously to a symbol (from anywhere). Symbol names are separated with a /. For instance, the symbol str is also present in a namespace called clojure.core and the corresponding full name is clojure.core/str The main purpose of symbols is to refer to things, i.e. to point to other values. When evaluating a program, symbols are looked up and replaced by their corresponding values.

more about symbols

(= str clojure.core/str)
;=> true

(name 'clojure.core/str)
;=> "str"

Keywords

Closely related to symbols and strings are keywords, which begin with a :.
Keywords are like strings in that they are made up of text, but are specifically intended for use as labels or identifiers. These are not labels in the sense of symbols, keywords are not replaced by any other value, they are just names, by themselves:

(type :cat)
;=> clojure.lang.Keyword

(str :cat)
;=> ":cat"

(name :cat)
;=> "cat"

Using keywords as map keys

since keywords are self-evaluating and provide fast equality checks, they are almost always used as map keys

Another important property of keywords, when used as map keys, is that they can be used as functions, taking a map as an argument to perform value lookups:

(def mouse-planet {:cats 180, :mice 9})   ;define a map
;=> #'user/mouse-planet

(:cats mouse-planet)                      ;lookup by keyword
;=> 180

(println (/ (:cats mouse-planet)          ;much more useful example
            (:mice mouse-planet))
         "cats per capita")
;=> 20 cats per capita

As enumerations

Since their value doesn't change, convenient keyword use case is enumeration. E.g. :mouse, :rat and :x-rat provide a nice visual delineation (for mouse types) within a source code.

there are other useful things we can do with keywords. We can use them As multimethod dispatch values and As directives, but we'll deal with that later

Qualifying your keywords

Keywords don't belong to any specific namespace, although it's a good practice to define them as if they do, because that way you provide a context:

user=> :not-in-ns
;=> :not-in-ns

user=> ::not-in-ns          ;fully qualified keyword
;=> :user/not-in-ns

user=> (ns another)
;=> nil

another=> :user/in-another  ;"fully qualified" keyword
;=> :user/in-another

another=> :haunted/name     ;namespace doesn't have to exist
;=> :haunted/name

double colon is used to fully qualify a keyword by prepending the current namespace name to the keyword name

equally named keywords are the same object in memory

Booleans

Only false and nil are logical false, all others values are true:

(boolean nil)
;=> false

(boolean 0)
;=> true

(boolean "hi")
;=> true

(boolean str)
;=> true

(and true false true)
;=> false

(and true true true)
;=> true

... and returns the first falsy value, or the last value if all are truthy:

(and 1 2 3)
;=> 3

(and -1 nil 2)
;=> nil

or returns the first truthy value, or the last value if all are falsy:

(or false 2 3)
;=> 2

(or false false nil)
;=> nil

not inverses the truthiness of the expression:

(not nil)
;=> true

Regular expressions

#"..." is Clojure’s way of writing a regular expression.
Clojure and Java have very similar Regex syntax.

(re-find #"cat" "mystic cat mouse")
;=> "cat"

(re-find #"cat" "only dogs here")
;=> nil

The parentheses, i.e. capturing group means that the regular expression should capture that part of the match. We get back a list containing the part of the string that matched the first parentheses, followed by the part that matched the second parentheses, etc:

(re-matches #"(.+):(.+)" "mouse:treat")
;=> ["mouse:treat" "mouse" "treat"]

;; capturing group in the regex causes each returned item to be a vector:
(re-seq #"\w*(\w)" "one-two/three")
;=> (["one" "e"] ["two" "o"] ["three" "e"])

Java's regex Pattern class has several methods that can be used directly, but only split is used regularly to split a string into an array of Strings:

(seq (.split #"," "1,2,3,4"))  ;this is how you call Java methods
;=> ("1" "2" "3" "4")

Java interoperability section.

The re-seq function returns a lazy sequence of all matches in a string:

(re-seq #"\w+" "one-two/three")
;=> ("one" "two" "three")

Java's regex engine includes a Matcher object which mutates in a non-thread-safe way as it walks through a string finding matches. This object is exposed in Clojure through the re-matcher function and can be used in combination with re-groups and re-find. It's recommended to avoid direct usage of all of these three functions.

Collections

A collection is a group of values. It’s a container which provides some structure, some framework, for the things that it holds. We refer to collection members as elements, or items.

All Clojure's collections support heterogeneous values (values of arbitrary types), together, in the same collection.

when a collection is evaluated (except lists), each of its contained items is evaluated first

Most of functions that work on collections aren't actually written to work on concrete collection types, but rather to work on abstract data types. For instance, some of abstractions in this space are:

  • Collection (common to all concrete types)
  • Sequential (ordered collections are lists and vectors)
  • Associative (maps associate keys with values, vectors associate indices with values)
  • Indexed (vectors, for example, can be quickly indexed into)

Lists

A PersistentList is a singly linked list where each node knows its distance from the end. List elements can only be found by starting with the first element and walking each prior node in order. List elements can only be added or removed from the left end.

In idiomatic Clojure code, lists are used almost exclusively to represent code forms, e.g. to call functions, macros, ... Code forms are then eval-ed or used as the return value for a macro.

Lists are rarely used for anything other than to represent Clojure source code, because they rarely offer any value over vectors

Literal lists are written with parentheses: (yankee hotel foxtrot).

When a list is evaluated, the first item of the list, yankee, will be resolved to a function, macro, or special form. If yankee is a function, the remaining items in the list will be evaluated in order, and the results will be passed to yankee as its parameters. If yankee is a macro or a special form, the remaining items in the list aren’t necessarily evaluated, but are processed as defined by the macro or operator.

special form is a form with special syntax or special evaluation rules that are typically not implemented using the base Clojure forms. An example of a special form is the . (dot) operator used for Java interoperability purposes.

(cons 1 (2 3))

;=> ClassCastException java.lang.Long cannot be cast to clojure.lang.IFn
;;which basically means that the number 2 cannot be used as a function

(cons 1 '(2 3))
;=> (1 2 3)

Remember, we quote lists (any everything else needed quoting) with a ' (or quote) to prevent them from being evaluated.

'(1 2 3)
;=> (1 2 3)

(type '(1 2 3))
;=> clojure.lang.PersistentList

There's also a syntax-quote (back-tick), which automatically qualifies all unqualified symbols in its argument:

`map
;=> clojure.core/map

`(map even? [1 2 3])
;=> (clojure.core/map clojure.core/even? [1 2 3])

You can also construct a list using list:

(list 1 2 3)
(1 2 3)
(= (list 1 2) (list 1 2))
true

You can modify a list by conjoining an element onto it (as always with lists, the new element goes to the beginning):

(conj (list 1 2 3) 4)
;=> (4 1 2 3)

(first (list 1 2 3))
;=> 1

(second (list 1 2 3))
;=> 2

(nth (list 1 2 3) 2)
;=> 3

Unlike some Lisps, the empty list in Clojure, (), isn't the same as nil. Lists are well-suited for small collections, or collections which are read in linear order, but are slow when you want to get arbitrary elements from later in the list.

Calling seq on a list returns the list itself, but more often, calling seq on a collection returns a new seq object for navigating that collection.

What lists aren't?

Probably the most common misuse of lists is to hold items that will be looked up by index. Though you can use nth to get the element, Clojure will have to walk the list from the beginning to find it.

Lists aren't queues. You can add items to one end, but you can't remove from the other.

PersistentQueue

Persistent immutable queue is a FIFO structure where conj adds to the rear, pop removes from the front and peek returns the front element without removal.

Clojure currently doesn't provide a core construction function for creating persistent queues, but there's a readily available empty queue instance to use, clojure.lang.PersistentQueue/EMPTY.

The printed representation for Clojure's queues isn't particularly informative, but we can change that by providing a method for them on the print-method:

(defmethod print-method clojure.lang.PersistentQueue
  [q, w]
  (print-method '<- w) (print-method (seq q) w) (print-method '-< w))

clojure.lang.PersistentQueue/EMPTY
;=> <-nil-<

Popping an empty queue results in just another empty queue. Peeking an empty queue returns nil

The mechanism for adding an element to a queue is conj:

(def tasks
  (conj clojure.lang.PersistentQueue/EMPTY
  :wake-up :shower :brush-teeth))
;=> #'mbo/tasks

tasks
;=> <-(:wake-up :shower :brush-teeth)-<

Clojure's persistent queue is implemented internally using two separate collections, the front being a seq and the rear being a vector. All insertions occur in the rear vector and all removals occur in the front seq, taking advantage of each collection's strength. When all the items from the front have been popped, the back vector is wrapped in a seq to become the new front and an empty vector is used as the new back

To get the front element, we use peek:

(peek tasks)
;=> :wake-up

To remove elements from the front of a queue, we use pop (although possible, it's non-idiomatic and suboptimal to use rest with queues, because it returns a seq, not a queue and screws up the speed guarantees):

(pop tasks)       ;; returns new list, remember?
;=> <-(:shower :brush-teeth)-<

(rest tasks)
;=> (:shower :brush-teeth)

Vectors

For fast access to every element, we use a vector. Vectors are surrounded by square brackets, just like lists are surrounded by parentheses. Because vectors aren’t evaluated like lists are, there is no need to quote the vector literal:

[1 2 3]
;=> [1 2 3]

(type [1 2 3])
;=> clojure.lang.PersistentVector

Vectors are similar to arrays, but are immutable and persistent

You can also create vectors with vector, or change other structures into vectors with vec:

(vector 1 2 3)
;=> [1 2 3]

(vec (list 1 2 3))
;=> [1 2 3]

conj on a vector adds to the end, not the start, like with lists:

(conj [1 2 3] 4)
;=> [1 2 3 4]

Our friends first, second, and nth work here too; but unlike lists, nth is fast on vectors. That’s because internally, vectors are represented as a very broad tree of elements, where each part of the tree branches into 32 smaller trees. Even very large vectors are only a few layers deep, which means getting to elements only takes a few hops, which is why retrieval operation from the interior of a vector takes essentially a constant time.

The important difference, when compared to lists, is that vectors evaluate each item in order. No function or macro is performed on a vector itself.

rest and next both return everything but the first element. They differ only by what happens when there are no remaining elements:

(rest [1 2 3 4])
;=> (2 3 4)

(next [1 2 3 4])
;=> (2 3 4)

rest returns logical true, next returns logical false. Each has their uses, but in almost every case they’re equivalent:

(next [1])
;=> nil

(rest [1])
;=> ()

last, surprisingly, returns the last element:

(last [1 2 3])
;=> 3

count, shockingly, returns element count:

(count [1 2 3 4])
;=> 4

You can use element's index to access it:

([1 2 3] 1)  ;vectors are in fact functions of their indices
;=> 2

Vectors and lists containing the same elements are considered equal:

(= [1 2] (list 1 2))
;=> true

If you already have a vector but want to "pour" several values into it, use into:

(let [m-vector [:a :b :c]]
  (into m-vector (range 10)))
;=> [:a :b :c 0 1 2 3 4 5 6 7 8 9]

;;if you want to return a vector the first arg to 'into' must be a vector
;;the second argument must be anything that works with 'seq' function

Primitive vectors

You can store primitive types inside of vectors using the vector-of function, which takes any of :int, :long, :float, :double, :byte, :short, :boolean and :char and returns an empty vector. This returned vector will act just like any other vector, except that, internally, it'll store its contents as primitives.
The new vector will try to coerce any additions into its internal type when they are being added:

(into (vector-of :int) [Math/PI 2 1.3])
;=> [3 2 1]

(into (vector-of :char) [100 101 102])
;=> [\d \e \f]

(into (vector-of :int) [1 2 609812734019519652839108477134])

;=> IllegalArgumentException Value out of range for long: 60981273401951...
;=> clojure.lang.RT.longCast (RT.java:1134)

Large vectors

Vectors are particularly efficient at three things (relative to lists):

  • adding or removing things from the right end of the collection
  • accessing or changing items in the interior of the collection by numeric index
  • walking in reverse order
(def a-to-e (vec (map char (range 65 70))))
;=> #'mbo/a-to-z

a-to-e
;=> [\A \B \C \D \E]

(nth a-to-e 4)   ;like with a map
;=> \E

(get a-to-e 4)   ;like with a map
;=> \E

(a-to-e 4)       ;invoking a vector as a function
;=> \E

vector-lookup-options

Since vectors are indexed, they can be efficiently walked in either direction. The seq and rseq return sequences that do exactly that:

(seq a-to-e)
;=> (\A \B \C \D \E)

(rseq a-to-e)
;=> (\E \D \C \B \A)

Any item in a vector can be "changed" using the assoc function (in a constant time). Clojure does this using structural sharing between the old and the new vectors, hence avoiding having to copy all the elements.

(assoc a-to-e 3 "former D")
;=> [\A \B \C "former D" \E]

The assoc function works only on indices that either already exist or, as a special case, exactly one step past the end of a vector (returned vector becomes one item larger).

Function replace uses assoc internally:

(replace {2 :a, 4 :b} [1 2 3 2 3 4])
;=> [1 :a 3 :a 3 :b]

The functions assoc-in, update-in and get-in are used to work with nested structures of vectors and/or maps. These functions take a series of indices to pick items from each more deeply nested level:

(def matrix
  [[1 2 3]
   [4 5 6]
   [7 8 9]])
;=> #'mbo/matrix

(get-in matrix [1 2])   ;second row, third column
;=> 6

(assoc-in matrix [1 2] 'x)
;=> [[1 2 3] [4 5 x] [7 8 9]]

update-in works the similar way, but instead of taking a value to "overwrite" an existing value, it takes a function to apply to an existing value. It'll replace the value at the given coordinates with the return value of the function:

(update-in matrix [1 2] * 100)
;=> [[1 2 3] [4 5 600] [7 8 9]]

;; the coordinates refer to the value 6 and the function is * taking an argument 100
;; so the slot becomes the return value of (* 6 100)

Remember? matrix itself never changed:

matrix
;=> [[1 2 3] [4 5 6] [7 8 9]]

Vectors as stacks

Clojure vectors uses conj as push and pop as pop to add and remove elements from the right side. Since vectors are immutable, pop returns a new vector with the rightmost element dropped, which is different from many mutable stack APIs, which generally return the dropped item. Consequently, peek becomes more important as the primary way to get an item from the top of the stack:

(peek m-stack)
;=> 3

(pop m-stack)
;=> [1 2]

(conj m-stack 4)
;=> [1 2 3 4]

(+ (peek m-stack) (peek (pop m-stack)))
;=> 5

If you're using a collection (any implementor of clojure.lang.IPersistentStack) as a stack, it's idiomatic to use conj instead of assoc, peek instead of last and pop instead of dissoc. That avoids unnecessary confusion about how the collection is being used

Lists also implement IPersistentStack, but the functions there operate on the left side, i.e. beginning of the list

Using vectors instead of reverse

When processing a list, it's pretty common to want to produce a new list in the same order. But if all you have is list, then you're left with backward list that needs to be reversed:

(defn m-map1 [f coll]
  (loop [coll coll, acc nil]
    (if (empty? coll)
      (reverse acc)
      (recur (next coll) (cons (f (first coll)) acc)))))
;=> #'mbo/m-map1

(m-map1 - (range 5))
;=> (0 -1 -2 -3 -4)

After the entire list has been walked once to produce the desired values, reverse walks it again to get them in the right order, which is inefficient and non-idiomatic. One way to avoid reverse is to use a vector as the accumulator, instead of a list:

(defn m-map2 [f coll]
  (loop [coll coll, acc []]
    (if (empty? coll)
      acc
      (recur (next coll) (conj acc (f (first coll)))))))
;=> #'mbo/m-map2

(m-map2 - (range 5))
;=> [0 -1 -2 -3 -4]

Subvectors

Provide a fast way to take a slice of an existing vector based on start and end indices. To produce a subvector we use the subvec function:

(subvec a-to-e 2 4)
;=> [\C \D]

;;first index is inclusive, but the second is exclusive

There's a special logic for taking a subvec of a subvec, in which case the newest subvector keeps a reference to the original vector, not the intermediate subvector, which keeps both the creation and use of sub-subvectors fast and efficient.

Vectors as MapEntries

Clojure's hash map, just like hash tables and dictionaries in many other languages, has a mechanism to iterate through the entire collection. Clojure's solution for this iterator is, you guessed it - seq.
Each item in this seq needs to include both the key and the value, hence they are wrapped in a MapEntry. When printed, each entry looks like a vector:

(first {:width 10 :height 20 :depth 15})
;=> [:depth 15]

;;not only does it look like a vector, it really is one:
(vector? (first {:width 10 :height 20 :depth 15}))
;=> true

;;which means you can use all the regular vector functions on it
;;it even supports destructuring*

destructuring?

;;e.g. the following locals, 'dimension' and 'amount', will take on the value of each
;;key/value pair in turn:
(doseq [[dimension amount] {:width 10, :height 20, :depth 15}]
  (println (str (name dimension) ":") amount " inches"))
;=> depth: 15  inches
;=> width: 10  inches
;=> height: 20  inches
;=> nil

(doc doseq)
;=> -------------------------
;=> clojure.core/doseq
;=> ([seq-exprs & body])
;=> Macro
;=>   Repeatedly executes body (presumably for side-effects) with
;=>   bindings and filtering as provided by "for".  Does not retain
;=>   the head of the sequence. Returns nil.

A MapEntry is its own type and has two functions for retrieving its contents, key and val, which do exactly the same thing as (nth m-map 0) and (nth m-map 1).

What vectors aren't?

Vectors are versatile, but there are some useful patterns where they might seem like a good fit, but in fact the aren't.

Vectors aren't sparse

If you have a vector of length n, the only position where you can insert a value is at index n, appending to the far right end. You can't skip some indices and insert at a higher index number (use hash map or sorted map).
Although you can replace values within a vector, you can't insert or delete items such that indices are adjusted (use finger trees for that).

Vectors aren't queues

Use a PersistentQueue for that.

Vectors aren't sets

If you want to find out whether a vector contains a particular value, you might be tempted to use the contains? function, but that function checks whether a specific key, not value is in a collection, which is really not useful for a vector.

Vectors are probably the most frequently used collection type in Clojure. They are used as literals for argument lists and let bindings, for holding large amounts of application data and as stacks and map entries

In almost all contexts, you can consider vectors, lists, and other sequences as interchangeable. They only differ in their performance characteristics, and in a few data-structure-specific operations.

Sets

Set is a collection of unordered, unique elements.

Sets are surrounded by #{...}:

#{1 2 3}
;=> #{1 2 3}

(#{:a :b :c :d} :c)
;=> :c

(#{:a :b :c :d} :e)
;=> nil

If you look at the previous two expressions, you'll see that the sets are in fact functions of their elements that return the matched element or nil if the element is not present (take a pause and read that a couple more times, examining the examples).

Elements can be accessed via the get function, which returns the queried value if it exists, or nil if it doesn't:

(get #{:a 1 2 :b} :b)
;=> :b

(get #{:a 1 2 :b} :k)
;=> nil

Searching a sequence for any of multiple items using a set and some

The some function takes a predicate and a sequence. It applies the predicate to each element in turn, returning the first truthy value returned by the predicate or else nil:

(some #{:b} [:a 1 :b 2])
;=> :b

(some #{1 :b} [:a 1 :b 2])
;=> 1

Using a set as the predicate supplied to some allows you to check whether any of the truthy values in the set are contained within the given sequence. This is a frequently used Clojure idiom for searching for containment within a sequence.

The key to understanding how sets determine whether an element is discrete lies in one simple statement: Given two elements, evaluating as equal, a set will contain only one, independent of concrete types:

(let s #{[1 2] (1 2)})  ;vector and list with the same items are considered equal

;=> IllegalArgumentException Duplicate key: (1 2)
;=> clojure.lang.PersistentHashSet.createWithCheck (PersistentHashSet.java:68)

Sorted sets

To ask for elements in a sorted order:

(sort #{2 4 1})
;=> (1 2 4)

As long as the arguments to the sorted-set function are mutually comparable, you'll receive a sorted set. Otherwise an exception is thrown:

(sorted-set :b :c :a)
;=> #{:a :b :c}

(sorted-set [3 4] [1 2])
;=> #{[1 2] [3 4]}

(sorted-set :b 2 :c)

;=> ClassCastException clojure.lang.Keyword cannot be cast to java.lang.Number
;=> clojure.lang.Util.compare (Util.java:152)

You can use your own comparator with sorted-set-by function.

Containment

contains? checks for existence of a key in a collection:

(contains? #{1 2 3} 3)
;=> true

Many newcomers to Clojure expect contains? to behave the same as Java's java.util.Collection#contains method, but it doesn't:

(contains? [1 2 4 3] 4)
;=> false

Since contains? looks for existence of a key it shouldn't work with sets, but it does. That is due to the fact that sets are implemented as maps with the same element as the key and value with an additional check for containment before insertion.

clojure.set namespace

clojure.set/intersection function works as you might expect. Given two sets, it returns a set of the common elements. Given n sets, it'll incrementally return the intersection of resulting sets and the next set:

(clojure.set/intersection #{:a :b :c} #{:d :c :b})
;=> #{:b :c}

(clojure.set/intersection #{:a :e :i :o :u}
                          #{:a :u :r}
                          #{:r :u :s})
;=> #{:u}

clojure.set/union works as expected.
clojure.set/difference "removes" all the elements from the first set that are also present in the second set (known as relative complement).

conj "adds" an element to the set:

(conj #{:a :b :c} :d)
;=> #{:a :c :b :d}

disj "removes" an element:

(disj #{1 2} 2)
;=> #{1}

You can make a set out of any other collection with set:

(set [2 5 1])
;=> #{1 2 5}

Maps

Map is an unsorted, associative key/value structure (associates keys with values). Maps are represented with curly braces {...} filled by alternating key/value pairs, with or without commas:

{:name "luka", :weight 3 :color "white"}   ;notice missing comma
;=> {:weight 3, :name "luka", :color "white"}

We can look up the value by key with get:

(get {"cat" "meow", "dog" "woof"} "cat")
;=> "meow"

get can also take a default value to return instead of nil, if the key doesn’t exist:

(get {:k :v} :w :default)
;=> :default

Maps can be used as verbs, directly:

({"a" 12, "b" 24} "b")  ;maps are functions of their keys
;=> 24

Keywords can also be used as verbs, which look themselves up:

(:raccoon {:weasel "queen", :raccoon "king"})
;=> "king"

In addition to literal syntax, hash maps can be created with the hash-map function:

(hash-map :a 1, :b 2, :c 3, :d 4, :e 5)
;=> {:a 1, :c 3, :b 2, :d 4, :e 5}

Maps support heterogeneous keys, which means that keys can be of any type and each key can be of a different type:

(let [m {:a 1, 1 :b, [1 2 3] "4 5 6"}]
  [(get m :a) (get m [1 2 3])])
;=> [1 "4 5 6"]

;;which can also be written without 'get', using map as a function of its keys:
(let [m {:a 1, 1 :b, [1 2 3] "4 5 6"}]
  [(m :a) (m [1 2 3])])
;=> [1 "4 5 6"]

Providing a map to a seq function returns a sequence of map entries:

(seq {:a 1, :b 2})
;=> ([:a 1] [:b 2])   ;returns key/value pairs contained in vectors

A new hash map can be created idiomatically using into:

(into {} [[:a 1] [:b 2]])
;=> {:a 1, :b 2}

;;even if pairs aren't vectors, they can easily be made to be
;;here, the map function, that has nothing to do with map-the-data-structure,
;;applies 'vec' to all the elements:
(into {} (map vec '[(:a 1) (:b 2)]))
;=> {:a 1, :b 2}

If the key/value pairs are laid out in a sequence consecutively, your pairs don't even have to be explicitly grouped. You can use apply to create a hash map:

(apply hash-map [:a 1 :b 2])
;=> {:a 1, :b 2}

Another idiomatic way to build a map is to use zipmap to "zip" together two sequences, the first of which contains keys and the second one values:

(zipmap [:a :b] [1 2])
;=> {:b 2, :a 1}

assoc "adds" an element to a map:

(assoc {:bolts 1088} :camshafts 3)
;=> {:camshafts 3, :bolts 1088}

assoc adds keys if they are not present, and replaces values if they are already there. If you associate a value onto nil, it creates a new map:

(assoc nil 5 2)
;=> {5 2}

merge yields a map containing all the elements of all given maps, preferring the values from later ones:

(merge {:a 1, :b 2} {:b 3, :c 4})
;=> {:c 4, :a 1, :b 3}

Remove map element with dissoc:

(dissoc {:a 1, :b 2, :c 4} :c)
;=> {:a 1, :b 2}

Sorted maps

The function sorted-map builds a map sorted by the comparison of its keys:

(sorted-map :b 1 :a 2 :c 0)
;=> {:a 2, :b 1, :c 0}

If you need alternative key ordering, or ordering for keys that are not naturally comparable, use sorted-map-by, which allows you to provide a comparison function:

(sorted-map-by #(compare %1 %2) "bac" 9 "abc" 2)
;=> {"abc" 2, "bac" 9}

;;Explanation:
;; by doing this:
#(compare %1 %2)   ;%1 stands for first parameter, in our case first map element

;; we declared an anonymous function inline, which is the same as writing:
(defn m-compare [first second]
  (compare first second))

;; and then passing that function to sorted-by-map:
(sorted-map-by m-compare "bac" 9 "abc" 2)
;=> {"abc" 2, "bac" 9}

Sorted maps (and sets) support efficient jump to a particular key and walk forward or backward from there through the collection. This is where subseq and rsubseq come in:

(def sm (sorted-map :a 1, :b 2, :c 3, :d 4, :e 5, :f 6, :g 7))

(subseq sm > :e)
;=> ([:f 6] [:g 7])

(subseq sm > :c < :f)
;=> ([:d 4] [:e 5])

There's one important difference in how sorted maps and sets handle numeric keys. A number can be represented by different types (long, int, float,...) and in a hash map, those types are preserved, while in a sorted map all those types are converted to the most precise one:

(assoc {1 :int} 1.0 :float)
;=> {1.0 :float, 1 :int}

(assoc (sorted-map 1 :int) 1.0 :float)  ;only one is kept, since maps are unique
;=> {1 :float}

When we're adding an element to a sorted map (or set), Clojure uses equality to determine both, the sort order and key presence (remember, maps have unique keys).

Array maps

Array maps are used to guarantee that the order of insertions will be preserved (just like vectors and arrays do):

(hash-map :a 1, :b 2, :c 3)
{:a 1, :c 3, :b 2}

(array-map :a 1, :b 2, :c 3)
{:a 1, :b 2, :c 3}

Like vectors, any item in a map literal is evaluated before the result is stored in the map. Unlike vectors, the order in which they are evaluated isn't guaranteed.

Persistence, immutability and functional programming

We'll start this section by giving a couple of phrases that encapsulate Clojure's immutability principles.

if two objects aren't equal forever, then they're technically never equal

there's a difference between a mutable object and a mutable reference. The default in Java is that there are references that may point to mutable data, but in Clojure, there are only mutable references

immutable objects are always thread safe (Brian Goetz)

Structural sharing

(def baselist (list :barnabas :adam))

(def lst1 (cons :willie baselist))

(def lst2 (cons :phoenix baselist))

baselist
;=> (:barnabas :adam)                 ;baselist stays unchanged

lst1
;=> (:willie :barnabas :adam)

lst2
;=> (:phoenix :barnabas :adam)

(= (next lst1) (next lst2))           ;they are not only equal
;=> true

(identical? (next lst1) (next lst2))  ;but also the same underlying object
;=> true

You can think of baselist as a common historical version of lst1 and lst2, but it's also the shared part of both lists. Not only are the next parts of both lists equal, they are identical (the same instance in memory).

Unlike lists, vectors and maps allow changes anywhere in the collection, not just on one end. This is made possible by the underlying data structure that those collection types are built upon - a tree.

A tree allows interior changes and still maintains shared structure between different versions/changes.

To demonstrate how structural sharing works, we'll build a simple tree, where each node will have three fields: a value, a left branch and a right branch:

{:val 50, :L nil, :R nil}

Our empty tree will be represented by nil and the map above will represent that empty tree after a single node (with value 50) has been added.

Let's start slowly. To handle just this initial case of adding a single node to an empty tree, we'll write xconj function like this:

(defn xconj [t v]     ;add item with value 'v' to tree 't'
  (cond
    (nil? t) {:val v, :L nil, :R nil})) ;if t is nil this will be the first node (root)

(xconj nil 50)
;=> {:val 50, :L nil, :R nil}

OK, that works, but let's not start sucking each others' dicks just yet (TMMr Wolf). We need to handle the case of adding an item to a non-empty tree.

So our tree doesn't even look like a tree yet:

{:val 50, :L nil, :R nil}

:L and :R are nil, so, for the sake of simplicity, we'll just make a rule that if an element does not point to some other element, it means it is nil. Presented this way, our tree looks like a piece of crap:

{:val 50, :L nil, :R nil}

When adding a node to a non-empty tree, in order to keep the tree sorted, we must follow a simple rule: the value of any node in our tree must be greater than its left child and smaller or equal to its right child.

So, in order to honor that rule, we need to compare the value being added first with the root, then with other nodes, down the tree, until we find the appropriate place for it.

Let's try adding a couple of elements. How would we add value 40 to our tree?

{:val 50, :L nil, :R nil}

Now that you saw this crucial image, take a look at the next one (let's keep our nils a bit longer, it will be easier to reason about the algorithm):

{:val 50, :L nil, :R nil}

How in earth did element 40 ended up there?

(defn xconj [t v]
  (cond
    (nil? t) {:val v, :L nil, :R nil}   ;does tree exist?
    (< v (:val t))              ;if v is smaller than the current item's value
          {:val (:val t),       ;make new item whose value is the value of current item
           :L (xconj (:L t) v), ;point :L to new tree by calling xconj on tree's :L child
           :R (:R t)}           ;point :R to current item's :R
    :else {:val (:val t)
           :L (:L t)
           :R (xconj (:R t) v)}
   ))


(defn xseq [t]
  (when t
    (concat (xseq (:L t)) [(:val t)] (xseq (:R t)))))

;;what is 'cond'?
(doc cond)
;=> -------------------------
;=> clojure.core/cond
;=> ([& clauses])
;=> Macro
;=>   Takes a set of test/expr pairs. It evaluates each test one at a
;=>   time.  If a test returns logical true, cond evaluates and returns
;=>   the value of the corresponding expr and doesn't evaluate any of the
;=>   other tests or exprs. (cond) returns nil.

This is how our tree looks like:

Unbalanced binary tree

This is how our tree would look like if it was balanced. Compare number of hops needed to find value 20.

Balanced binary tree

persistent collections are immutable, in-memory (not on-disk) collections that allow you to preserve historical versions of their state.

Since arrays are mutable, any changes happen in-place:

(def ds (into-array [:frane :luka :glupaca]))  ;mutable
;=> #'mbo/ds

;; btw Glupaca is the name of our cat (which means something like "stupid women")

(seq ds)     ;only for nice REPL printout
;=> (:frane :luka :glupaca)

(aset ds 2 :suzi)   ;replace third element of the array
;=> :suzi

(seq ds)
;=> (:frane :luka :suzi)

Using one of Clojure's persistent data structures shows the difference:

mbo=> (def ds [:frane :luka :suzi])  ;init persistent collection
;=> #'mbo/ds

ds            ;print ds to REPL
;=> [:frane :luka :suzi]

(def ds8 (replace {:suzi :glupaca} ds))   ;replace third element and bind the
;=> #'mbo/ds8                             ;new collection to ds8

ds            ;ds did not change
;=> [:frane :luka :suzi]

ds8           ;the newly created collection
;=> [:frane :luka :glupaca]

Symbols

Closest thing to a variable in Clojure. Primarily used to provide a name for a given value, i.e. to refer to function parameters, local variables, globals, and Java classes.

Unlike keywords, symbols are not unique based solely on their names:

(identical? 'node 'node)
;=> false

;; identical? returns true only if symbols are the same object
(let [x 'node y x] (identical? x y))
;=> true

Each node symbol is a discrete object that only happens to share a name. Though name is the basis for symbol equality:

(= 'node 'node)
;=> true

(name 'node)
;=> "node"

We can define a meaning for a symbol within a specific expression, using let. The let expression first takes a vector of bindings: alternating symbols and values that those symbols are bound to within the remainder of the expression.

“Let the symbol cats be 5, and construct a string composed of "I have ", cats, and " cats":

(let [cats 5] (str "I have " cats " cats."))
;=> "I have 5 cats."

Let bindings, also called locals, apply only within the let expression itself. They also override any existing definitions for symbols at that point in the program. For example, we can redefine addition to mean subtraction, for the duration of a let:

(let [+ -] (+ 2 3))
;=> -1

That definition does not apply outside the let:

(+ 2 3)
;=> 5

We can also provide multiple bindings. Since Clojure doesn’t care about spacing, alignment, or newlines, we’ll write this on multiple lines, for clarity:

(let [person "joseph"
      num-cats 186]
  (str person " has " num-cats " cats!"))  ;the body
;=> "joseph has 186 cats!"

When multiple bindings are given, they are evaluated in order. Later bindings can use previous bindings:

(let [cats 3
      legs (* 4 cats)]
  (str legs " legs all together"))
;=> "12 legs all together"

a symbol whose name is prefixed with a namespace, followed by a slash, is called fully qualified symbol:

clojure.core/map
;=> #<core$map clojure.core$map@2a0406c4>

clojure.set/union
;=> #<set$union clojure.set$union@1be2bcc8>

The body is sometimes described as an implicit do (see blocks bellow) because it follows the same rules: you may include any number of expressions and all will be evaluated, but only the value of the last one is returned.

Because they’re immutable, locals can’t be used to accumulate results. Instead, you'd use a high level function or loop/recur form.

To summarize, let defines the meaning of symbols within an expression. When Clojure evaluates a let, it replaces all occurrences of those symbols in the rest of the let expression with their corresponding values, then evaluates the rest of the expression.

Metadata

Clojure allows the attachment of metadata to various objects, including symbols. with-meta function takes an object and a map and returns another object of the same type with the metadata attached:

(let [x (with-meta 'node {:js true})    ;attach :js to 'node and assign to x
      y (with-meta 'node {:js false})]
  [(= x y)            ;true because they both hold the same symbol, 'node
   (identical? x y)   ;false because they are different instances
   (meta x)
   (meta y)])
;=> [true false {:js true} {:js false}]

Symbols and namespaces

Like keywords, symbols don't belong to any specific namespace:

user=> (ns what-where)
;=> nil

what-where=> (def one-simbol 'where-is-it)
;=> #'what-where/one-simbol

what-where=> one-simbol
;=> where-is-it

what-where=> (resolve 'one-simbol)
;=> #'what-where/one-simbol  ;looks like namespace-qualified symbol
                             ;but it's just a characteristic of symbol evaluation

what-where=> `one-symbol   ;back tick
;=> what-where/one-symbol

Functions

Functions are a first-class type in Clojure. They can be used the same as any value (stored in Vars, held in collections, passed as arguments and returned as a result of other functions).

(let [x] (+ x 1))

We can’t actually evaluate this program, because there’s no value for x yet. It could be 1, or 4, or 1453. We say that x is unbound, because it has no binding to a particular value. This is the nature of the function: an expression with unbound symbols.

Function definition:

(fn [x] (+ x 1))
;=> #<user$eval1487$fn__1488 user$eval1487$fn__1488@6b7d28db>

Named function definition:

(let [twice (fn [x] (* 2 x))]
   (+ (twice 1)
   (twice 3)))
;=> 8

let bindings describe a similar relationship, but with a specific set of values for those arguments. let is evaluated immediately, whereas fn is evaluated later, when bindings are provided.

Vars

Once a let is defined, there’s no way to change it. If we want to redefine symbols for everyone, even code that we didn’t write, we need a new construct, a mutable variable.

(def cats 5)
;=> #'user/cats

(type #'user/cats)
;=> clojure.lang.Var

cats
;=> 5

def defines a type of value we haven’t seen before: a Var. Vars, like symbols, are references to other values. When evaluated, a Var is replaced by its corresponding value.

def also binds the symbol cats (and its globally qualified equivalent user/cats) to that Var.

The symbol inc points to the Var #'inc, which in turn points to the function #<core$inc clojure.core$inc@16bc0b3c>.

We can see the intermediate Var with resolve:

'inc     ;symbol
;=> inc

(resolve 'inc)
;=> #'clojure.core/inc    ;variable

(eval 'inc)
;=> #<core$inc clojure.core$inc@d6206b5>  ;value

Why those two levels of indirection? Unlike with symbol, we can change the meaning of a Var for everyone, globally, at any time.

Vars don't require a value. Instead we can just declare them and, by doing so, defer the binding of value.

(def y)
;=> #'user/y

;; if we try to use it:
y

;=> java.lang.IllegalStateException: Var user/y is unbound

Named functions

(def half 
  (fn [number] (/ number 2)))
;=> #'user/half

(half 8)
;=> 4

Creating a function and binding it to a variable is so common that it has its own form: defn, which is a macro that is short for def fn:

(defn half [number] (/ number 2))
;=> #'user/half

(half 8)
;=> 4

Functions don’t have to take an argument. We’ve seen functions which take zero arguments, like (+):

(defn half [] 1/2)
;=> #'user/half

But if we try to use our earlier form with one argument, Clojure complains that the arity (the number of arguments to a function) is incorrect:

(half 8)

;=> ArityException Wrong number of args (1) passed to: user$half
;=> clojure.lang.AFn.throwArity (AFn.java:437)

To handle multiple arities, functions have an alternate form, instead of an argument vector and a body, one provides a series of lists, each of which starts with an argument vector, followed by the body:

(defn half
   ([] 1/2)
   ([x] (/ x 2)))
;=> #'user/half

(half)
;=> 1/2

(half 8)
;=> 4

Multiple arguments work just like you expect. Just specify an argument vector of two, or three, or however many arguments the function takes.
Some functions can take any number of arguments. For that, Clojure provides &, which slurps up all remaining arguments as a list:

(defn vargs [x y & more-args]
   {:x x
    :y y
    :more more-args})
;=> #'user/vargs

(vargs 1 2)
;=> {:x 1, :y 2, :more nil}

(vargs 1 2 3 4 5)
;=> {:x 1, :y 2, :more (3 4 5)}

x and y are mandatory, though there don’t have to be any remaining arguments.

To keep track of what arguments a function takes, why the function exists, and what it does, we usually include a docstring. Docstrings help fill in the missing context around functions, to explain their assumptions, context, and purpose to the world:

(defn launch
   "Launches a spacecraft into the given orbit by initiating a
    controlled on-axis burn. Does not automatically stage, but
    does vector thrust, if the craft supports it."
   [craft target-orbit]
   "OK, we don't know how to control spacecraft yet.")
;=> #'user/launch

Docstrings are used to automatically generate documentation for a Clojure programs, but you can also access them from the REPL (The user=> prompt refers to the top-level namespace of the default REPL):

(doc launch)
;=> -------------------------
;=> user/launch
;=> ([craft target-orbit])
;=>   Launches a spacecraft into the given orbit by initiating a
;=>    controlled on-axis burn. Does not automatically stage, but
;=>    does vector thrust, if the craft supports it.
;=> nil

doc tells us the full name of the function, the arguments it accepts, and its docstring. This information comes from the launch Var’s metadata, and is saved there by defn. We can inspect metadata directly with the meta function:

(meta #'launch)
;=> {:arglists ([craft target-orbit]), :ns #<Namespace user>, :name launch, :column 1,
;=> :doc "Launches a spacecraft into the given orbit by initiating a\n   controlled 
;=> on-axis burn. Does not automatically stage, but\n   does vector thrust, if the 
;=> craft supports it.", :line 1, :file "/tmp/form-init523009510157887861.clj"}

There’s some other juicy information in there, like the file the function was defined in and which line and column it started at, but that’s not particularly useful since we’re in the REPL, not a file. However, this does hint at a way to answer our motivating question: how does the type function work?

Blocks

When you have a series or block of expressions that need to be treated as one, use do. All the expressions will be evaluated, but only the last one will be returned:

(do
   6
   (+ 5 4)
   3)
;=> 3

The expressions 6 and (+ 5 4) are perfectly legal. The addition in (+ 5 4) is even done, but the value is thrown away, only the final expression 3 is returned

How does type work?

type, like all functions, is a kind of object with its own unique type:

type
;=> #<core$type clojure.core$type@2761df2a>

(type type)
;=> clojure.core$type

This tells us that type is a particular instance, at memory address 39bda9b9, of the type clojure.core$type. clojure.core is a namespace which defines the fundamentals of the Clojure language, and $type tells us that it’s named type in that namespace. None of this is particularly helpful, though. Maybe we can find out more about the clojure.core$type by asking what its supertypes are:

(supers (type type))
;=> #{java.io.Serializable java.lang.Runnable clojure.lang.AFunction 
;=> clojure.lang.IMeta clojure.lang.AFn java.lang.Object clojure.lang.IObj 
;=> java.util.Comparator clojure.lang.Fn java.util.concurrent.Callable 
;=> clojure.lang.IFn}

This is a set of all the types that include type. We say that type is an instance of clojure.lang.AFunction, or that it implements or extends java.util.concurrent.Callable, and so on. Since it’s a member of clojure.lang.IMeta it has metadata, and since it’s a member of clojure.lang.AFn, it’s a function. Just to double check, let’s confirm that type is indeed a function:

(fn? type)
;=> true

type can take a single argument, which it calls x. If it has :type metadata, that’s what it returns. Otherwise, it returns the class of x. Let’s take a deeper look at type’s metadata for more clues:

(doc type)
;=> -------------------------
;=> clojure.core/type
;=> ([x])
;=>   Returns the :type metadata of x, or its Class if none
;=> nil

This function was first added to Clojure in version 1.0, and is defined in the file clojure/core.clj, on line 3109:

(meta #'type)
;=> {:ns #<Namespace clojure.core>, :name type, :arglists ([x]), :column 1, 
;=> :added "1.0", :static true, :doc "Returns the :type metadata of x, or its Class 
;=> if none", :line 3109, :file "clojure/core.clj"}

We could go dig up the Clojure source code and read its definition there, or we could ask Clojure to do it for us. Aha! Here, at last, is how type works. It’s a function which takes a single argument x, and returns either :type from its metadata, or (class x).

(source type)
;=> (defn type 
;=>   "Returns the :type metadata of x, or its Class if none"
;=>   {:added "1.0"
;=>    :static true}
;=>   [x]
;=>   (or (get (meta x) :type) (class x)))
;=> nil

Sequences

A sequential collection is one that holds a series of values without reordering them.

A sequence is a sequential collection that represents a series of values that may or may not exist yet (may have concrete values, may be lazy or empty). Few composite types are actually sequences, though several such as vectors are sequential. All an object needs to do to be a sequence is to support two core functions, first and rest

A seq is a simple API for navigating collections which consists of two functions, first and rest

If two sequentials have the same values in the same order, = will return true for them, even if their concrete types are different:

(= [1 2 3] '(1 2 3))
;=> true

Conversely, even if two collections have the same values in the same order, if one is a sequential collection and the other isn't, = will return false:

(= [1 2 3] #{1 2 3})
;=> false

If the collection is empty, seq always returns nil and never an empty sequence (that goes for all other functions that return a seq, e.g. next).

Clojure classifies each composite data type into 3 partitions: sequentials, maps and sets. Everything that implements java.util.List is included in the sequential partition. Generally things that fall into other two partitions include set or map in their name

Every collection type provides at least one kind of seq object for walking through its elements

Recursion

cons, makes a list beginning with the first argument, followed by all the elements in the second argument:

(cons 1 [2 3 4])
;=> (1 2 3 4)

Problem of incrementing all elements of a vector:

(defn inc-first [nums]
   (if (first nums)
     ; If there's a first num, build a new list with cons
     (cons (inc (first nums))
           (rest nums))
     ; If there's no first num, return an empty list
     (list)))
;=> #'user/inc-first

(inc-first [])
;=> ()

(inc-first [1 2 3])
;=> (2 2 3)

What if we called our function on rest?

(defn inc-all [nums]
   (if (first nums)
     (cons (inc (first nums))
           (inc-all (rest nums)))
     (list)))
;=> #'user/inc-all

(inc-all [1 2 3 4])
;=> (2 3 4 5)

This technique is called recursion, and it is a fundamental principle in working with collections, sequences, trees, graphs or any problem which has small parts linked together. There are two key elements in a recursive program:

  • Some part of the problem which has a known solution
  • A relationship which connects one part of the problem to the next

Incrementing the elements of an empty list returns the empty list. This is our base case, the ground to build on. Our inductive case, also called the recurrence relation, is how we broke the problem up into incrementing the first number in the sequence, and incrementing all the numbers in the rest of the sequence. The if expression bound these two cases together into a single function, a function defined in terms of itself. Let’s parameterize our inc-all function to use any transformation of its elements:

(defn transform-all [f xs]
   (if (first xs)
     (cons (f (first xs))
           (transform-all f (rest xs)))
     (list)))
;=> #'user/transform-all

(transform-all inc [1 2 3 4])
;=> (2 3 4 5)

Loop

When using recursion, you sometimes want to loop back not to the top of the function, but to somewhere inside the function body.
The loop acts exactly like let but provides a target for recur to jump to:

(defn sum-down-from [initial-x]
   (loop [sum 0, x initial-x]
     (if (pos? x)
       (recur (+ sum x) (dec x))
       sum)))
;=> #'user/sum-down-from

Upon entering the loop, the locals sum and x are initialized (like in let). A recur always loops back to the closest enclosing loop or fn. The loop locals are rebound to the values given in recur.
recur works only from the tail position.

keyword transforms a string to keyword:

(transform-all keyword ["aa" "bb" "cc"])
;=> (:aa :bb :cc)

To wrap every element in a list:

(transform-all list ["aa" "bb" "cc"])
;=> (("aa") ("bb") ("cc"))

We basically implemented map function:

(map inc [1 2 3 4])
;=> (2 3 4 5)

The function map relates one sequence to another. The type map relates keys to values. There is a deep symmetry between the two: maps are usually sparse, and the relationships between keys and values may be arbitrarily complex. The map function, on the other hand, usually expresses the same type of relationship, applied to a series of elements in a fixed order.

Clojure has a special form called recur that's specifically used for tail recursion:

(defn print-down-from [x]
   (when (pos? x)  ;return when x is no longer positive
     (println x)
     (recur (dec x))))
;=> #'user/print-down-from

(print-down-from 5)
;=> 5
;=> 4
;=> 3
;=> 2
;=> 1
;=> nil

when is same as if, except it doesn't have the else part and it provides an implicit do in order to perform side-effects

This is nearly identical to how you’d structure a while loop in an imperative language.One significant difference is that the value of x isn’t decremented somewhere in the body of the loop. Instead, a new value is calculated as a parameter to recur, which immediately does two things: rebinds x to the new value and returns control to the top of print-down-from.
If the function has multiple arguments, the recur call must as well, just as if you were calling the function by name instead of using the recur special form. And just as with a function call, the expressions in the recur are evaluated in order first and only then bound to the function arguments simultaneously.

Building Sequences

We can use recursion to expand a single value into a sequence of values, each related by some function. For instance (pos? returns true if num is greater than zero, else false):

(defn expand [f x count]
   (if (pos? count)
     (cons x (expand f (f x) (dec count)))))
;=> #'user/expand

(expand inc 0 10)
;=> (0 1 2 3 4 5 6 7 8 9)

Our base case is x itself, followed by the sequence beginning with (f x). That sequence in turn expands to (f (f x)), and then to (f (f (f x))), and so on. Each time we call expand, we count down by one using dec. Once the count is zero, if returns nil, and evaluation stops.

Clojure has a more general form of this function, called iterate:

(take 10 (iterate inc 0))
;=> (0 1 2 3 4 5 6 7 8 9)

Since this sequence is infinitely long, we’re using take to select only the first 10 elements. We can construct more complex sequences by using more complex functions:

(take 10 (iterate (fn [x] (if (odd? x) (+ 1 x) (/ x 2))) 10))
;=> (10 5 6 3 4 2 1 2 1 2)

repeat constructs a sequence where every element is the same:

(take 10 (repeat "a"))
;=> ("a" "a" "a" "a" "a" "a" "a" "a" "a" "a")

(repeat 5 "b")
;=> ("b" "b" "b" "b" "b")

repeatedly simply calls a function (f) to generate an infinite sequence of values, over and over again, without any relationship between elements. For an infinite sequence of random numbers:

(rand)
;=> 0.6934524557647231

(rand)
;=> 0.1355414232605504

(take 3 (repeatedly rand))
;=> (0.18806021884865332 0.5231673860825672 0.38244349544358525)

range generates a sequence of numbers between two points. (range n) gives n successive integers starting at 0. (range n m) returns integers from n to m-1. (range n m step) returns integers from n to m, separated by step:

(range 5)
;=> (0 1 2 3 4)

(range 5 8)
;=> (5 6 7)

(range 5 25 5)
;=> (5 10 15 20)

cycle returns an infinite lazy sequence of repetitions of the items in a collection:

(take 6 (cycle (range 5 50 5)))
;=> (5 10 15 20 25 30)

Transforming Sequences

map applies a function to each element, but it has a few more tricks up its sleeve:

(map (fn [n vehicle] (str "I've got " n " " vehicle "s"))
   [0 200 9]
   ["car" "train" "kiteboard"])
;=> ("I've got 0 cars" "I've got 200 trains" "I've got 9 kiteboards")

If given multiple sequences, map calls its function with one element from each sequence in turn. So the first value will be (f 0 "car"), the second (f 200 "train"), and so on. Like a zipper, map folds together corresponding elements from multiple collections. To sum three vectors, column-wise:

(map + [1 2 3]
        [4 5 6]
        [3 2 1])
;=> (8 9 10)

If one sequence is bigger than another, map stops at the end of the smaller one. We can exploit this to combine finite and infinite sequences. For example, to number the elements in a vector:

(map (fn [index element] (str index ". " element))
      (iterate inc 0)
      ["erlang" "scala" "haskell"])
;=> ("0. erlang" "1. scala" "2. haskell")

Transforming elements together with their indices is so common that Clojure has a special function for it: map-indexed:

(map-indexed (fn [index element] (str index ". " element))
              ["erlang" "scala" "haskell"])
;=> ("0. erlang" "1. scala" "2. haskell")

You can also tack one sequence onto the end of another, like so:

(concat [1 2 3] [:a :b :c] [4 5 6])
;=> (1 2 3 :a :b :c 4 5 6)

Another way to combine two sequences is to riffle them together, using interleave:

(interleave [:a :b :c] [1 2 3])
;=> (:a 1 :b 2 :c 3)

And if you want to insert a specific element between each successive pair in a sequence, try interpose:

(interpose :and [1 2 3 4])
;=> (1 :and 2 :and 3 :and 4)

To reverse a sequence, use ... you guessed it, reverse:

(reverse [1 2 3])
;=> (3 2 1)

(reverse "woolf")
;=> (\f \l \o \o \w)  ;not a string

Strings are sequences too! Each element of a string is a character, written \f. You can rejoin those characters into a string with apply str:

(apply str (reverse "woolf"))
;=> "floow"

…and break strings up into sequences of chars with seq:

(seq "sato")
;=> (\s \a \t \o)

To randomize the order of a sequence, use shuffle:

(shuffle [1 2 3 4 5])
;=> [4 3 5 1 2]

(apply str (shuffle (seq "abracadabra")))
;=> "raradbabaac"

Subsequences

take selects the first n elements drop removes the first n elements take-last and drop-last operate on the last n elements:

(take 3 (range 10))
;=> (0 1 2)

(drop 3 (range 10))
;=> (3 4 5 6 7 8 9)

(take-last 3 (range 10))
;=> (7 8 9)

(drop-last 3 (range 10))
;=> (0 1 2 3 4 5 6)

take-while and drop-while work just like take and drop, but use a function to decide when to stop:

(take-while pos? [3 2 1 0 -1 -2 10])
;=> (3 2 1)

In general, one can cut a sequence in twain by using split-at with a particular index. There’s also split-with, which uses a function to decide when to cut:

(split-at 4 (range 10))
;=> [(0 1 2 3) (4 5 6 7 8 9)]

(split-with number? [1 2 3 :mark 4 5 6 :mark 7])
;=> [(1 2 3) (:mark 4 5 6 :mark 7)]

Notice that because indexes start at zero, sequence functions tend to have predictable number of elements.(split-at 4) yields four elements in the first collection, and ensures the second collection begins at index four. (range 10) has ten elements, corresponding to the first ten indices in a sequence. (range 3 5) has two (as in 5 - 3 = 2) elements. These choices simplify the definition of recursive functions as well. We can select particular elements from a sequence by applying a function. To pull up all positive numbers in a list, use filter:

(filter pos? [1 5 -4 -7 3 0])
;=> (1 5 3)

filter looks at each element in turn, and includes it in the resulting sequence only if (f element) returns a truthy value. Its complement is remove, which only includes those elements where (f element) is false or nil:

(remove string? [1 "tur" :apple])
;=> (1 :apple)

One can group a sequence into chunks using partition, partition-all, or partition-by. For instance, one might group alternating values into pairs like this:

(partition 2 [:cats 5 :bats 27 :crocs 0])
;=> ((:cats 5) (:bats 27) (:crocs 0))

Separate a series of numbers into negative and positive subsequences:

(partition-by neg? [1 2 3 2 1 -1 -2 -3 -2 -1 1 2])
;=> ((1 2 3 2 1) (-1 -2 -3 -2 -1) (1 2))

partition-all n collection may include partitions with fewer than n items at the end:

(partition-all 3 [1 2 -5 3 2 1 -1 -2 -3 -2 -1 1 2])
;=> ((1 2 -5) (3 2 1) (-1 -2 -3) ( -2 -1 1) ( 2))

while partition may not:

(partition 3 [1 2 -5 3 2 1 -1 -2 -3 -2 -1 1 2])
;=> ((1 2 -5) (3 2 1) (-1 -2 -3) (-2 -1 1))

Collapsing subsequences

After transforming a sequence, we often want to collapse it in some way, in order to derive some smaller value. For instance, we might want the number of times each element appears in a sequence:

(frequencies [:meow :mrrrow :meow :meow])
;=> {:meow 3, :mrrrow 1}

To group elements by some function:

(pprint (group-by :first [{:first "Li"    :last "Zhou"}
                           {:first "Sarah" :last "Lee"}
                           {:first "Sarah" :last "Dunn"}
                           {:first "Li"    :last "O'Toole"}]))
;=> {"Li" [{:last "Zhou", :first "Li"} {:last "O'Toole", :first "Li"}],
;=>  "Sarah" [{:last "Lee", :first "Sarah"} {:last "Dunn", :first "Sarah"}]}

Here we’ve taken a sequence of people with first and last names, and used the :first keyword (which can act as a function!) to look up those first names. group-by used that function to produce a map of first names to lists of people, kind of like an index. In general, we want to combine elements together in some way, using a function. Where map treated each element independently, reducing a sequence requires that we bring some information along. The most general way to collapse a sequence is reduce:

(doc reduce)
;=> -------------------------
;=> clojure.core/reduce
;=> ([f coll] [f val coll])
;=>   f should be a function of 2 arguments. If val is not supplied,
;=>   returns the result of applying f to the first 2 items in coll, then
;=>   applying f to that result and the 3rd item, etc. If coll contains no
;=>   items, f must accept no arguments as well, and reduce returns the
;=>   result of calling f with no arguments.  If coll has only 1 item, it
;=>   is returned and f is not called.  If val is supplied, returns the
;=>   result of applying f to val and the first item in coll, then
;=>   applying f to that result and the 2nd item, etc. If coll contains no
;=>   items, returns val and f is not called.

(reduce + [1 2 3 4])
;=> 10

(reduce + 1 [1 2 3 4])
;=> 11

To see the reducing process in action, we can use reductions, which returns a sequence of all the intermediate states:

(reductions + [1 2 3 4])
;=> (1 3 6 10)

Oftentimes we include a default state to start with. For instance, we could start with an empty set, and add each element to it as we go along:

(reduce conj #{} [:a :b :b :b :a :c])
;=> #{:a :c :b}

Reducing elements into a collection has its own name: into. We can conj [key value] vectors into a map, for instance, or build up a list:

(into {} [[:a 2] [:b 3]])
;=> {:a 2, :b 3}

(into (list) [1 2 3 4])
;=> (4 3 2 1)

Because elements added to a list appear at the beginning, not the end, this expression reverses the sequence. Vectors conj onto the end, so to emit the elements in order, using reduce, we might try:

(reduce conj [] [1 2 3 4 5])
;=> [1 2 3 4 5]

Remember?

(conj [-1 0] [1 2 3 4 5])
;=> [-1 0 [1 2 3 4 5]]               ; not [-1 0 1 2 3 4 5]

This looks like a map function. All that’s missing is some kind of transformation applied to each element:

(defn my-map [f coll]
   (reduce (fn [output element]
             (conj output (f element)))
           []
           coll))
;=> #'user/my-map

(my-map inc [1 2 3 4])
;=> [2 3 4 5]

So map is just a special kind of reduce. What about, say, take-while?

(defn my-take-while [f coll]
   (reduce (fn [out elem]
             (if (f elem)
               (conj out elem)
               (reduced out)))
           []
           coll))
;=> #'user/my-take-while

We’re using a special function here, reduced, to indicate that we’ve completed our reduction early and can skip the rest of the sequence.

(my-take-while pos? [2 1 0 -1 0 1 2])
;=> [2 1]

Most of Clojure’s sequence functions are lazy. For instance, we can increment every number from zero to infinity:

(def infseq (map inc (iterate inc 0)))
;=> #'user/infseq

(realized? infseq)
;=> false

That function returned immediately. Because it hasn’t done any work yet, we say the sequence is unrealized. It doesn’t increment any numbers at all until we ask for them:

(take 10 infseq)
;=> (1 2 3 4 5 6 7 8 9 10)

(realized? infseq)
;=> true

Lazy sequences also remember their contents, once evaluated, for faster access.

Find the sum of the products of consecutive pairs of the first 1000 odd integers.

(reduce +
         (take 1000
               (map (fn [pair] (* (first pair) (second pair)))
                    (partition 2 1
                      (filter odd?
                        (iterate inc 0))))))
;=> 1335333000

Homework:

  1. Write a function to find out if a string is a palindrome.
(defn palindrome? [word] 
  (== 0 (compare word (apply str (reverse word)))))
  1. Find the number of cs in “abracadabra”.
(defn occurs-count [c word]
  (get (frequencies (seq word)) c 0))
  1. Write your own version of filter.
;=> TBD

Java interop

Clojure is symbiotic with its host, providing its rich and powerful features, while Java provides an object model, libraries and runtime support.

Clojure strings are Java Strings, numbers are Numbers, collections implement Collection, fns implement Callable and Runnable, ...

Core abstractions, such as seq, are Java interfaces (ISeq).
Clojure seq library works on Java Iterables, Strings and arrays.
It is possible to implement and extend Java interfaces and classes.

Accessing static class members

... is trivial:

java.util.Locale/JAPAN
;=> #<Locale ja_JP>

Idiomatic Clojure prefers that you access static class members using a syntax like accessing a namespace-qualified Var:

(Math/sqrt 9)  ; the same as (java.lang.Math/sqrt 9)
;=> 3.0

Creating Java class instances

(new java.util.HashMap {"foo" 42 "bar" 8 "baz" "beep boop"})
;=> {"baz" "beep boop", "foo" 42, "bar" 8}

;;idiomatic:
(java.util.HashMap. {"foo" 42 "bar" 8 "baz" "beep boop"})   ;dot = constructor call
;=> {"baz" "beep boop", "foo" 42, "bar" 8}

Accessing Java instance members with the dot operator

To access instance properties, precede the property or method name with a dot:

(.x (java.awt.Point. 10 20))  ;create a new Point and access its member x
;=> 10

To access instance methods, the dot form allows an additional argument to be passed to the method:

(.divide (java.math.BigDecimal. "42") 2K)
;=> 21M

; what's this M for?
(type 2M)
;=> java.math.BigDecimal

Setting Java instance properties

In the absence of mutator methods, in the form of setX, Java instance properties can be set using set! function:

(let [origin (java.awt.Point. 0 0)]
  (set! (.x origin) 15)     ;(set! <instance_member_access> <value>)
  (str origin))
;=> "java.awt.Point[x=15,y=0]"

The .. macro

For now, think of a macro as a convenience function. We'll meet them later.
In Java, it's a common practice to chain together a sequence of method calls, e.g:

new java.util.Date().toString().endsWith("2014")  ;java

... which is equivalent to this Clojure expression:

(.endsWith (.toString (java.util.Date.)) "2014")
;=> true

Well, anyone would agree that this is rather difficult to read.
To remedy this, Clojure provides us with .. macro:

(.. (java.util.Date.) toString (endsWith "2014"))
;=> true

The doto macro

In Java, it's also common to initialize a fresh instance by calling a set of mutators:

java.util.HashMap props = new java.util.HashMap();  /* java */
props.put("HOME", "/home/myself");
props.put("SRC",  "src");
props.put("BIN",  "classes");

That's obviously overly verbose, but it can be streamlined using the doto macro:

(doto (java.util.HashMap.)       ;do to HashMap all these things
  (.put "HOME" "/home/myself")
  (.put "SRC"  "src")
  (.put "BIN"  "classes"))
;=> {"HOME" "/home/myself", "BIN" "classes", "SRC" "src"}

Exceptions

Like Java, Clojure also provides try, catch, finally and throw forms:

(throw (Exception. "Thrown"))

;=> Exception Thrown  user/eval1201 (form-init8547084957850583270.clj:1)

(defn throw-catch [f]
  [(try
    (f)
    (catch ArithmeticException e "You did not? Not by zero! Noooooooooooo...")
    (catch Exception e (str "You blew it " (.getMessage e)))
    (finally (println "returning... ")))])
;=> #'user/throw-catch

(throw-catch #(/ 10 5))
;=> returning... 
;=> [2]

(throw-catch #(/ 10 0))
;=> returning... 
;=> ["You did not? Not by zero! Noooooooooooo..."]

(throw-catch #(throw (Exception. "dawg!")))
;=> returning... 
;=> ["You blew it dawg!"]

Clojure doesn't adhere to checked exception requirements, like Java does.

When an exception is thrown in REPL, the result is stored in a Var named *e, which allows you to get more detail about the expression, such as the stack trace:

(.printStackTrace *e)

Namespaces

Provide a way to bundle related functions, macros and values.

Creating namespaces using ns macro

(ns mbo.core.strings)
;=> nil
;=> mbo.core.strings=>    ;from now on, this is our REPL prompt

There's also a Var *ns*, which holds the value of the current namespace.
We know from before that any Var created will be a member of the current namespace.

mbo.core.strings=> (defn report-ns [] (str "The current namespace is " *ns*))
;=> #'mbo.core.strings/report-ns

mbo.core.strings=> (report-ns)
;=> "The current namespace is mbo.core.strings"

mbo.core.strings=> (defn sing [] (println "Marjane Marjane, Marjane Marjane, ..."))
;=> #'mbo.core.strings/sing

mbo.core.strings=> sing    ;Clojure looks it up in the current namespace
;=> #<strings$sing mbo.core.strings$sing@3987b05>

mbo.core.strings=> (ns mbo.core.compat)   ;create another namespace
;=> nil

mbo.core.compat=> (report-ns)             ;try invoking function from another ns

mbo.core.compat=> CompilerException java.lang.RuntimeException: 
;=> Unable to resolve symbol: report-ns in this context, 
;=> compiling:(/tmp/form-init8547084957850583270.clj:1:1)

mbo.core.compat=> (mbo.core.strings/report-ns) ;fully qualified name works as expected
;=> "The current namespace is mbo.core.compat"

referring to a namespace symbol using fully qualified name will only work for namespaces created locally or those previously loaded. Read on, it'll become clear...

Using :require directive to load other namespaces

mbo.core.compat=> (ns mbo.core.set
                    (:require clojure.set))
;=> nil
;=> mbo.core.set=>  ;changed to new ns

mbo.core.set=> (clojure.set/intersection #{1 2 3} #{2 3 4})
;=> #{2 3}

mbo.core.set=> (intersection #{1 2 3} #{2 3 4})  ;invoke a clojure.set function directly

;=> CompilerException java.lang.RuntimeException: Unable to resolve symbol: 
;=> intersection in this context, compiling:(/tmp/form-init8547084957850583270.clj:1:1)

This construct indicates that we want the clojure.set namespace loaded, but we don't want the mappings of that namespace's symbols to mbo.core.set functions.

We can also use :as directive to create an additional alias to clojure.set:

mbo.core.set=> (ns mbo.core.set-alias
                    #_=>   (:require [clojure.set :as s]))
;=> nil

mbo.core.set-alias=> (s/intersection #{1 2 3} #{2 3 4})
;=> #{2 3}

The qualified namespace form (e.g. clojure.set) looks the same as a call to a static class method. The difference is that a namespace symbol can only be used as a qualifier, whereas a class symbol can also be referenced independently:

mbo.core.set-alias=> clojure.set

;=> mbo.core.set-alias=> CompilerException java.lang.ClassNotFoundException: 
;=> clojure.set, compiling:(/tmp/form-init8547084957850583270.clj:1:691) 

mbo.core.set-alias=> java.lang.Object
;=> java.lang.Object

That vagaries of namespace mappings from symbols to Vars, both qualified and unqualified, have the potential for confusion between class names and static methods. In the beginning, that is. The differences will begin to feel natural as we progress (at least that's what The joy of Clojure book promises :) One of the Clojure idioms is to use my.Class and my.ns for naming classes and namespaces, to help eliminate potential confusion.

Loading and creating mappings with :use

:use, unlike :require, maps Vars in another namespace to names in your own. That is typically used to avoid calling each function or macro with the qualifying namespace symbol:

mbo.core.set-alias=> (ns mbo.test
                #_=>   (:use [clojure.string :only [capitalize]]))
;=> nil

mbo.test=> (map capitalize ["one" "two"])
;=> ("One" "Two")
;=> mbo.test=>

:only is used to indicate that only the listed functions should be mapped in the new namespace (good practice). The :exclude directive does the opposite.
:use, besides creating mappings, implicitly invokes :require.

the idiomatic strategy for avoiding conflicts is to use :require with :as to create a namespace alias

Create mappings with :refer

:refer is a directive that works almost exactly like :use, except that it only creates mappings for libraries that have already been loaded (by being previously defined, by being one of Clojure's core namespaces or by having been explicitly loaded using :require).

Loading Java classes with :import

mbo.test=> (ns mbo.java
      #_=>   (:import [java.util HashMap]
      #_=>            [java.util.concurrent.atomic.AtomicLong]))
;=> nil

mbo.java=> (HashMap. {"happy?" true})
;=> {"happy?" true}

Any classes in the java.lang package are implicitly imported when namespaces are created

nil punning

Since, in Clojure, everything except nil and false is true, empty collections evaluate to true in boolean context. We need a way to test whether a collection is empty or not.

This is where nil punning comes in:

(seq [1 2 3])
;=> (1 2 3)

(seq [])
;=> nil

seq function returns a sequence view of a collection or nil if the collection is empty.

(defn print-seq [s]
  (when (seq s)
    (prn (first s))     ;prn prints each object in a newline (to the output stream)
    (recur (rest s))))

(print-seq [1 2])
;=> 1
;=> 2
;=> nil

The use of seq as a terminating condition is the idiomatic way of testing whether a sequence is empty

Destructuring

Allows you to place a collection of names in a binding form where normally you'd put just a single name.

Destructuring is loosely related to pattern matching (found in Haskell or Scala), but much more limited in scope. For full-featured pattern matching in Clojure use matchure.

Perhaps the simplest form of destructuring is picking apart a sequential thing (e.g. a vector or a list), giving each item a name:

(let [[fname mname lname] ["Frane" "Luka" "Bonaci"]]
  (str lname ", " fname " " mname))
;=> "Bonaci, Frane Luka"

;; although this is syntactically correct, it isn't factually accurate
;; my month and a half old twins are named Frane and Luka :)
;; what can I say, my wife is a hero (that's what everybody's telling me these days)
;; hmmm, I wonder why?

This was a so called positional destructuring, which, as you might expect, doesn't work on maps and sets, because they are not logically aligned sequentially. But it does work on java.util.regex.Matcher and anything implementing CharSequence and java.util.RandomAccess interfaces.

Destructuring with a vector

We can also use an ampersand in a destructuring vector to indicate that any remaining values of the input should be collected into a (possibly lazy) seq:

(let [[a b c & more] (range 10)]
  (println "a b c are: " a b c)
  (println "the rest is: " more))
;=> a b c are:  0 1 2
;=> the rest is:  (3 4 5 6 7 8 9)

A useful feature of vector destructuring is :as, which is used to bind a local to the entire collection. It must be placed at the end, even after the & local (if it exists):

(let [range-vec (vec (range 10))
  [a b c & more :as all] range-vec]
  (println "a b c are: " a b c)
  (println "the rest is " more)
  (println "all is: " all))
;=> a b c are:  0 1 2
;=> the rest is  (3 4 5 6 7 8 9)
;=> all is:  [0 1 2 3 4 5 6 7 8 9]

Notice the difference between & and :as. While :all produces a vector, & results with a seq.

Destructuring with a map

(def full-name-map
  {:fname "Frane" :mname "Luka" :lname "Bonaci"})

(let [{fname :fname, mname :mname, lname :lname} full-name-map]
  (str lname ", " fname " " mname))
;=> "Bonaci, Frane Luka"

Here, the :keys feature might come in handy:

(let [{:keys [fname mname lname]} full-name-map]
  (str lname ", " fname " " mname))
;=> "Bonaci, Frane Luka"

By using :keys, we're telling Clojure that the next form will be a vector of names that it should convert to keywords (e.g. :fname), in order to look