Functional Programming

in JavaJava technology seminar

SECR, 2007Andrei Solntsev

Purpose

Seminar gives an overview of

Functional Programming methodsand

its applications in Java for

coding Business Logicand

its customization

Agenda

FP overview Sandwich algorithm FP main features Code samples Business Logic with FP FP libraries for Java

FP Overview

Computationis executing statements to change the program state.

Imperative programming

Functional programming

Computationis evaluation of expressions

The focus is on what, not how

Expressions are formed by using functions to combine basic values

Program consists of a sequence of commands.

Sandwich algorithm

Function createSandwich

1. Take a bread2. Spread bread with butter3. Put cheese on the bread4. return result

Imperative

return put ( cheese,spread(butter, bread) )

Functional

Sandwich algorithm

If we want to use sausage instead of cheese ?

Let’s pass sausage/cheese as input parameter

No problems!

Sandwich algorithm

1. Take a lower2. Spread lower with middle3. Put upper on the middle4. return result

Function createSandwich (lower, middle, upper)

return put ( upper,spread(middle, lower) )

Function createSandwich (lower, middle, upper)

No problems!

bread butter sausage

Sandwich algorithm

If we want to put butter instead of spreading ?

Imperative programming: Problem!

Functional programming: not a problem

Sandwich algorithm

1. Take a lower2. if mode = ‘put’

put middle on lower else

spread middle on lower end if

3. Put upper on the middle4. return result

Procedure createSandwich (lower, middle, upper, mode)

Imperative programming: Problem!

bread butter sausage put

Alternative: create 2 different functions Code duplication

Sandwich algorithm

return put ( upper,action(middle, lower) )

Function createSandwich (lower, middle, upper, action)

Functional programming: not a problem

bread butter sausage put

Action is a function with 2 parameters

• spread• put• …

createSandwich is a higher-order function which takesanother function as a parameter

FP main features

What is Functional Programming?

• Closures and higher order functions

• Lazy evaluation

• Recursion as a mechanism for control flow

• Enforcement of referential transparency

• No side-effects

FP Languages

• Lisp (AutoCad)

• Haskell, Scheme, Logo

• XSLT

Where a traditional imperative program might use a loop to traverse a list, a functional style would often use a higher-order function, map, that takes as arguments a function and a list, applies the function to each element of the list, and returns a list of the results.

Code Samples in Haskell

add :: Integer -> Integer -> Integeradd x y =  x + y

functions

inc :: Integer -> Integerinc = add 1

map :: (a->b) -> [a] -> [b]map f  []       =  []map f (x:xs)    =  f x : map f xs

zip (x:xs) (y:ys)  = (x,y) : zip xs yszip  xs     ys     = []

Uncurriedfunction

Function can bereturned as a value !

Higher-orderfunction

curriedfunction

Code Samples in Haskell

ones = 1 : ones

Infinite data structures

numsFrom n = n : numsFrom (n+1)

squares = map (^2) (numsfrom 0)

take 5 squares => [0,1,4,9,16]

Code Samples in Haskell

Fibonacci sequence

fib = 1 : 1 : [ a+b | (a,b) <- zip fib (tail fib) ]

FP-Style code example in Java

java.util.Properties

Properties properties = new Properties();properties.setProperty(“firstName", groom.getFirstName());properties.setProperty(“lastName", groom.getLastName());properties.setProperty(“salary", groom.getSalary());return parameters;

return

Imperative

Functional

return new Properties() .setProperty(“firstName", groom.getFirstName()) .setProperty(“lastName", groom.getLastName()) .setProperty(“salary", groom.getSalary());• Pros

• Cons

FP-Style code example In Java

StringBuffer

StringBuffer sb = new StringBuffer();sb.append(“a”);sb.append(“b”);sb.append(“c”);return sb.toString();

return new StringBuffer() .append(“a”); .append(“b”); .append(“c”) .toString();

Imperative

Functional

• Pros • Cons ?

FP: Pros and Cons

Pros

• Reliable code• Readable• Reusable• …

• Non-natural for human• Non-natural for computer• Performance

Cons

Example: Quick Sort algorithm

Code sample: Quicksort

Quicksort in Haskell

qsort [] = []

qsort (x:xs) = qsort elts_lt_x ++

[x] ++

qsort elts_greq_x

where

elts_lt_x = [y | y <- xs, y < x]

elts_greq_x = [y | y <- xs, y >= x]

Code sample: Quicksort

qsort( a, lo, hi ) int a[], hi, lo;{

int h, l, p, t;if (lo < hi) {

l = lo; h = hi; p = a[hi];do {

while ((l < h) && (a[l] <= p))l = l+1;

while ((h > l) && (a[h] >= p))h = h-1;

if (l < h) {t = a[l]; a[l] = a[h]; a[h] = t;

}} while (l < h);

t = a[l]; a[l] = a[hi]; a[hi] = t;qsort( a, lo, l-1 );qsort( a, l+1, hi );

}}

Quicksort in C

FP: Pros and Cons

Pros

• Reliable code• Readable• Reusable• …

• Non-natural for human• Non-natural for computer• Performance

Cons

Example: Quick Sort algorithm

In Java, FP suits for implementingBusiness Logic

Programs are easier to design, write and maintain, butprogrammer has less control over the machine.

Business logic with FP

GroomFilter

List suitableGrooms = new ArrayList();

for (groom in allGrooms){ if (minAge > -1 && groom.getAge() < minAge)

continue;

if (maxAge > -1 && groom.getAge() > maxAge)continue;

suitableGrooms.add(groom);}

return suitableGrooms;

List filterGrooms(List allGrooms, int minAge, int maxAge)

If age is -1 thenDon’t check age

Business logic with FP

GroomFilter

List suitableGrooms = new ArrayList();

for (groom in allGrooms){ if (groomChecker.accept(groom))

suitableGrooms.add(groom);}

return suitableGrooms;

List filterGrooms(List allGrooms, Filter groomChecker)

Pass functionas parameter

Business logic with FP

public interface Filter { /** * Method defines whether given object is accepted. * @param obj any Object * @return true iff object is accepted */ boolean accept(Object obj);}

Business logic with FP

public interface Filter { boolean accept(Object obj);

public static final Filter ACCEPT = new Filter() {

public boolean accept(Object obj){return true;} };

public static final Filter NOT_NULL = new Filter() {

public boolean accept(Object obj){return obj!=null;} };

public static final Filter NEGATE ..; public static final Filter IS_NULL = …;}

Predefinedvalues

Business logic with FP

Client 1

List suitableGrooms grooms = GroomFilter.filterGrooms(…, new Filter() {

public boolean accept(Object obj){ return ((Groom) obj).getAge() > 23;}

});

Client 2

List suitableGrooms = GroomFilter.filterGrooms(…,Filter.ACCEPT);

Closure – object representing a function

Anonymous classesare often used as closures

25th frame

25th frame

Parameterized Closures

StringFilter

public class StringFilter implements Filter{ public static startsWith(final String prefix) {

return new Filter{ public boolean accept(Object o){

return ((String) o).startsWith(prefix); }};

}

public static endsWith (final String postfix) {…} public static contains (final String substring) {…} public static matches (final String regexp) {…}};

Composition of functions

Composition of functions: AND

public class AND implements Filter{ public AND (Filter filter1, Filter filter2) {

this.filter1 = filter1;this.filter2 = filter2;

}

public boolean accept(Object obj) {

return filter1.accept(obj) && filter2.accept(obj);

}};

FP Applications: Filters

FilteredIterator

public class FilteredIterator implements Iterator{ public FilteredIterator(Iterator iterator, Filter filter);}

CollectionsUtils

static List collectList(Iterator it);static Set collectSet(Iterator it);static List filterList (List original, Filter filter);static Set filterSet (Set originalSet, Filter filter);

FP Applications: Filters

Given: a list of all grooms’ names.Goal: find all names with prefix “Mr.”

List gentlemen = new LinkedList();

for (Iterator it = groomsNames.iterator(); it.hasNext(); ){ String name = (String) it.next(); if (name != null &&

name.startsWith(“Mr.”)) {

gentlemen.add(name); }}

return gentlemen;

Imperative

FP Applications: Filters

Functional

return CollectionsUtils.filterList(allGrooms,StringFilter.startsWith(“Mr.”));

Given: a list of all grooms’ names.Goal: find all names with prefix “Mr.”

Transformer

public interface Transformer{ Object transform(Object sourceObject);}

ListTransformer

public class ListTransformer{ public List transform(List sourceList, Transformer transformer);}

FP Applications: Transformers

FP Applications: Transformers

Given: list of GroomsGoal: create list grooms’ names

List groomsNames = new ArrayList();

for (Iterator it = allGrooms.iterator(); it.hasNext(); ){ Groom groom = (Groom) it.next(); groomsNames.add(groom.getName());}

return groomsNames;

Imperative

FP Applications: Transformers

return ListTransformer.transform( allGrooms,

new Transformer(){ public Object transform(Object obj) {

return ((Groom) obj).getName(); }});

Functional

Given: list of GroomsGoal: create list grooms’ names

Business Logic customization

Example using Plexus container

import org.codehaus.plexus.embed.Embedder;

public List findSuitableGrooms(Client woman){

Filter clientGroomFilter = (Filter) embedder.lookup( “groomFilter”, woman.getName());

return GroomFilter.filterGrooms(allGrooms, clientGroomFilter);

}

META-INF/plexus/components.xml

<component-set> <components> <component> <role>groomFilter</role> <role-hint>default</role-hint> <implementation>examples.Filter.ACCEPT</implementation> </component>

<component> <role>groomFilter</role> <role-hint>Maril Strip</role-hint> <implementation>examples.filters.OlderThan25</implementation> </component>

<component> <role>groomFilter</role> <role-hint>Jenifer Lopez</role-hint> <implementation>examples.filters.SalaryBiggerThan10000</implementation> </component>

</components></component-set>

Business Logic customization

Conclusion

I hope this article has provided you with a good foundation for incorporating closures and higher order functions into your Java code, as well as giving you a glimpse of the beauty and effectiveness of functional programming.

FP Libraries for Java

Commons Functors: Function Objects for Javahttp://jakarta.apache.org/commons/sandbox/functor

JGA: Generic Algorithms for Java http://jga.sourceforge.net

http://plexus.codehaus.org

Articles

Functional programming in the Java languagehttp://www-128.ibm.com/developerworks/library-combined/j-fp.html

Use recursion effectively in XSLhttp://www-128.ibm.com/developerworks/xml/library/x-xslrecur

Why Functional Programming Mattershttp://www.math.chalmers.se/~rjmh/Papers/whyfp.html

Introduction to Haskellhttp://www.haskell.org/tutorial/