Glossary¶

A glossary of terminology specific to Clojure. Terms are listed in alphabetical order.

What about...? Are we missing a term you you need? Let us know! You can ask by filing a new issue.

NOTE: portions of this glossary have been adapted from the Clojure Doc Site under the CC-BY-SA 3.0 license.

arity¶

The number of arguments a function takes is its arity. If it’s written to take a variable number of args, it’s referred to as variadic.

Functions can have multiple arity (for example, a function might have 2 different bodies: one for when 2 args are passed, and another when 3 args are passed).

binding-form¶

Could mean one of two things:

the expression you’re binding to in a let-binding. It might be a simple name, or it might be a data structure used for destructuring.
Clojure provides the binding macro, used for setting the thread-local value of a dynamic var. The whole expression (form) is sometimes referred to as the “binding form”.

classpath¶

The search path used by the JVM to locate classes which are not part of the Java standard class library. May include jar files.

comparator¶

A function that takes two args and compares them. Returns -1, 0, or 1 depending whether the first arg is less than, equal to or greater than the second. The stock comparator that Clojure.core comes with is compare.

coordinates¶

The “group-id/artifact-id version-string” identifier used in your project.clj to indicate a particular dependency.

destructuring¶

The handy trick used in a let-binding to “unpack” the values from a data structure into the locals you’re going to use. See also binding-form.

dereference¶

To get the value of a reference type. You can use the deref function for this, or else some syntactic sugar: @some-ref-type.

entry¶

A key/value pair in a map. Try (type (first {:a 1 :b 2})) and see that it returns clojure.lang.MapEntry.

evaluator¶

todo

form¶

A valid s-expression. For example: (+ 1 1) and (defn foo [x] (* x x)).

head retention¶

Lazy sequences are still persistent. If you make another data structure using one, the original lazy sequence will be kept around and not garbage-collected. If the lazy sequence in infinite, and grows very large, it can cause performance problems or even an out-of-memory error. Accidentally keeping around a lazy sequence like this is referred to as “head retention”.

homoiconicity¶

Where the code and the data is represented by the same structure. This allows the code to be treated as data, and the data to be treated as code. This feature of Clojure, and other Lisps, allows for macros in the language, since they can operate on code as a data structure, and to return a transformation of that structure to be the representation of new code.

idempotent¶

An operation that when given the same inputs will produce the same result when called one or more times. An idempotent function may produce a side effect, such a updating a ref or an atom, but will only produce the side effect once. An idempotent function is different than a pure function, in that a pure function will produce no side effects.

identity¶

A logical entity in your program that may change over time — it may take on different states at different times, but it still means the same logical entity. Clojure uses reference types to represent identities. This is not to be confused with the identity function that just returns the argument given to it.

implicit do¶

The bodies of some expressions act like do in that you can include multiple expressions in them, and the expressions will be evaluated in the order they appear, with the resulting value of the body being the last expression evaluated. Forms that do this include: when, when-let, fn, defn, let, loop, and try.

intern¶

A method of storing values or immutable data structures as a single copy of the item, allowing for more space-efficiency, and possibly time-efficiency, with the trade off of requiring more time being required when interning the item. When the string “clojure” is interned, all instances of the string “clojure” will reference the exact same instance, instead of having multiple string objects with the same value of “clojure”.

keyword¶

A Clojure scalar data type whose literal syntax looks :like :this. They are like numbers and strings in that they evaluate to themselves, and are most often seen being used as keys in hash-maps.

lazy¶

Clojure can (and often does) create sequences for you that aren’t fully computed. Upon casual inspection they look just like a regular list, but particular values in them are only computed the moment you ask for them — not sooner.

This has the added benefit that you can easily create infinite sequences that don’t consume infinite memory.

Many of the built-in Clojure functions return lazy sequences.

let-binding¶

AKA, “binding vector”, or just “bindings”: in a let (and expressions that work like let, for example, defn, loop, loop, & fn), the vector that comes first where you specify lexical bindings.

libspec¶

todo

macro¶

A special type of function which is transforms a S-Expression read in and applies a transformation to the S-Expression resulting in a new form. This process is called macro-expansion, and is done as part of the Clojure reader.

map¶

Either refers to the built in map function, or else means “a hash-map object”.

memoization¶

The ability to cache a result of a function call by given arguments, and return the result without having to do the calculation again. Memoization is a time-space trade off in that more memory is used to store the results of a function call to be able to return the value instead of having to keep spending time doing the calculation involved in the function.

metadata¶

An extra map that you can attach to a collection value (or a symbol), which contains data about the data you’re attaching it to. Use meta to see the metadata of a given value.

namespaced keyword¶

When you put two colons in front of a keyword’s name — for example ::foo — it is a so-called “namespaced keyword”, and is expanded by the reader to become :current-namespace/foo.

nullipotent¶

An operation with no side effects. The result of calling the function one or more times is the same as if it was never called. Queries are typically good examples of functions that are nullipotent, as they do not modify the state of the object or structure they are queried against.

predicate¶

A function taking one or more args and returning a boolean (true or false). Its name typically ends with a question mark. Some examples: nil?, zero?, string?.

pure function¶

A function that given the same inputs will always produce the same result. A pure function also does not have any observable side effects and cannot depend on any outside state, other than that which was given as arguments to the function. A pure function’s result also cannot change during the execution of the program or between executions of the program, as the dependency on outside state can lead to changes in the result of the function. Pure functions are also referentially transparent.

reader¶

todo

reader macro¶

Syntax that the Clojure reader recognizes as special syntactic sugar, for example, #"", #{}, quoting, etc.

realize¶

When the next value in a lazy sequence is accessed for the first time, and is computed so as to made available, it is said to have been “realized”. This term is also used to refer to the status of promises, futures, and delays. That is, if a promise (for example) is realized then that means its value has been delivered and is accessible via dereferencing.

reference types¶

Vars, atoms, refs, and agents are all reference types. They are mutable in the sense that you can change to what value they refer, and Clojure provides thread-safe mechanisms for doing so.

referential transparency¶

An expression that will always return the same result for the values given, and can be substituted for the resulting value, without effecting the program. The advantage of referential transparent expressions is that they can be memoized, and be the subject of various compilier optimizations.

reify¶

todo

REPL¶

Short for: “Read, Eval, Print, Loop”. The REPL reads in text through the reader transforming it into a Clojure data structure, evaluates the data structure as code, prints the result of the evaluation, and loops back waiting to read the next input string.

rest args¶

The extra args passed to a variadic function, for example if my-func were defined like (defn my-func [a b & more] ...), then called like (my-func 1 2 3 4 5), then 3, 4, & 5 are the “rest args”.

s-expression¶

Short for Symbolic Expression. A S-Expression is a data structure able to represent both simple datastructes such as literals, or complex data structures such as nested expressions. Due to their versatile nature, S-Expressions are able to represent both data in Clojure, as well as the Clojure code itself, allowing Clojure to be a homoiconic language.

state¶

The value that a given identity may have at a given time. When you change the state of an identity, you’re changing to which value it refers. Clojure uses values to represent states.

STM (Software Transactional Memory)¶

Software Transactional Memory (STM) is a concurrency control method to coordinate and control access to shared storage as an alternative to lock-based synchronization. Clojure’s STM uses multiversion concurrency control (MVCC) as an alternative to lock-based transactions, as well as ensuring changes are made atomically, consistently, and in isolation. It does this by taking a snapshot of the ref, making the changes in isolation to the snapshot, and apply the result. If the STM detects that another transaction has made an update to the ref, the current transaction will be forced to retry.

symbol¶

An identifier that refers to vars or local values.

tagged literals¶

(Formerly called “reader literals”.)

Some literals begin with a hash mark “#” (so-called “dispatch macros”); for example, #{} for sets and #"" for regex literals. Starting with Clojure 1.4, you can create your own #-prefixed literal which causes the reader to parse the form following it using a function or macro of your own choosing/devising. It’s in this way that you can tag a literal to be handled specially by the reader.

For more info, see the “Tagged Literals” section of the reader doc.

threading macros¶

The thread-first (->) and thread-last (->>) macros. “Threading” refers to how they pass values to each subsequent argument in the macro, not concurrency.

thrush¶

A combinator. Not the same thing as the thread-first macro. More info at http://blog.fogus.me/2010/09/28/thrush-in-clojure-redux/ if you’re curious.

transaction¶

todo

type erasure¶

Java-related: Java generics allow you to specify a type for a collection. This way you don’t have to cast every object you pull out of an ArrayList like in the old days. This is a courtesy of the java compiler. The java runtime doesn’t know about generics — the compiler does all the checking for you, then the type information is discarded at runtime. In Clojure, this discarding is referred to as type erasure.

value¶

An immutable object, such as the number 1, the character \a, the string “hello”, or the vector [1 2 3]. In Clojure, all scalars and built-in core data structures are values.

variadic¶

A function that can take a variable number of arguments. See also rest args.