Scribd
Upload
216 views

Functional Programming

Functional programming emphasizes evaluating expressions rather than executing commands. It treats functions as first-class objects that can be passed into other functions as values.

Uploaded by

Tushar Shrimali
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
216 views

Functional Programming

Functional programming emphasizes evaluating expressions rather than executing commands. It treats functions as first-class objects that can be passed into other functions as values.

Uploaded by

Tushar Shrimali
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 37

PRINCIPLES OF PROGRAMMING

LANGUAGES
CS F301 – Functional Programming
PROF. KUNAL KORGAONKAR
BOOKS
• Ravi Sethi , "Programming Languages: Concepts and Constructs" 2nd Edition by
Addison Wesley
• Robert W. Sebesta, "Concepts of Programming Languages", 2nd Edition by The
Benjamin/Cummings, Publishing Company, Inc
• Aho, Lam, Sethi and Ullman, "Compilers Principles, Techniques, and Tools". Pearson
Education. Low Price Edition. 2004
• Michael L. Scott, Programming Language Pragmatics, Fourth Edition, Morgan
Kaufmann Publishers
• Benjamin C. Pierce, Types and Programming Languages, January 2002, The MIT
Press
• Shriram Krishnamurthi, Programming Languages: Application and Interpretation,
Creative Commons License
• Erich Gamma, Richard Helm, Ralph Johnson, John Vlissides, Design Patterns:
Elements of Reusable Object-Oriented Software, Addison-Wesley
• Ivan Čukić, Functional Programming in C++, Manning Books (available online)
References
Contents based on and an interpretation of:
Ivan Čukić, Functional Programming in C++,
Manning Books (available online).
Introduction to functional
programming
• What is functional programming?
• Functional programming is a style of programming
that emphasizes the evaluation of expressions,
rather than execution of commands.
• The expressions in these languages are formed by
using functions to combine basic values.
• Examples: Haskell and Lisp
• Relationship with object-oriented programming
• We shall look at a concrete example of imperative
vs. declarative programming
Introduction to functional
programming
• Broadly speaking, FP is a style of programming in
which the main program building blocks are functions
as opposed to objects and procedures.
• A program written in the functional style doesn’t
specify the commands that should be performed to
achieve the result, but rather defines what the result is.
Introduction to functional
programming
• Consider a small example: calculating the sum of a
list of numbers. In the imperative world, you
implement this by iterating over the list and adding
the numbers to the accumulator variable.
• You explain the step-by-step process of how to sum
a list of numbers.
• On the other hand, in the functional style, you need
to define only what a sum of a list of numbers is.
The computer knows what to do when it’s required
to calculate a sum.
Introduction to functional
programming
• One way you can do this is to say that the sum of a
list of numbers equals the first element of the list
added to the sum of the rest of the list, and that the
sum is zero if the list is empty.
Introduction to functional
programming
• This difference is the origin of the
terms imperative and declarative programming.
• Imperative means you command the computer to do
something by explicitly stating each step it needs to
perform in order to calculate the result.
• Declarative means you state what should be done, and
the programming language has the task of figuring out
how to do it.
• You define what a sum of a list of numbers is, and the
language uses that definition to calculate the sum of a
given list of numbers.
Relationship with object-
oriented programming
• The object-oriented paradigm is based on creating
abstractions for data.
• It allows the programmer to hide the inner
representation inside an object and provide only a
view of it to the rest of the world via the object’s
API.
• The FP style creates abstractions on the functions.
• This lets you create more-complex control
structures than the underlying language provides.
The program input is a list of files. The program needs to
return the number of newlines in each file as its output.
Example – Steps
1. Open each file.
2. Define a counter to store the number of lines.
3. Read the file one character at a time, and increase
the counter every time the newline character (\n)
occurs.
4. At the end of a file, store the number of lines
calculated.
std::vector<int>
count_lines_in_files(const std::vector<std::string>& files)
{
std::vector<int> results;
char c = 0;
for (const auto& file : files) {
int line_count = 0;
Calculating std::ifstream in(file);
the number of while (in.get(c)) {
lines the if (c == '\n') {
imperative line_count++;
way
}
}
results.push_back(line_count);
}
return results;
}
int count_lines(const std::string& filename)
{
std::ifstream in(filename);
return std::count(
std::istreambuf_iterator<char>(in),
std::istreambuf_iterator<char>(),
'\n');
}
Using
std::vector<int>
std::count to
count newline count_lines_in_files(const std::vector<std::string>& files)
characters {
std::vector<int> results;
for (const auto& file : files) {
results.push_back(count_lines(file));
}
return results;
}
Std:: count
std::count() returns number of occurrences of an
element in a given range. Returns the number of
elements in the range [first,last) that compare equal to
val.

// Returns count of occurrences of value in


// range [begin, end]
int count(Iterator first, Iterator last, T &val)
std::vector<int>
count_lines_in_files(const std::vector<std::string>& files)
{
Mapping files
std::vector<int> results(files.size());
to line counts
by using std::transform(files.cbegin(), files.cend(),
std::transform results.begin(),
count_lines);
return results;
}
Std::transform

Applies an operation sequentially to the elements of one (1) or two (2) ranges and
stores the result in the range that begins at result.

(1) unary operation: Applies op to each of the elements in the


range [first1,last1) and stores the value returned by each operation in the range
that begins at result.
(2) binary operation : Calls binary_op using each of the elements in the
range [first1,last1) as first argument, and the respective argument in the range that
begins at first2 as second argument. The value returned by each call is stored in
the range that begins at result.
std::vector<int>
count_lines_in_files(const std::vector<std::string>& files)
{
Transformation return files | transform(count_lines);
using ranges
}
Pure functions
• One of the most significant sources of software
bugs is the program state.
• It’s difficult to keep track of all possible states a
program can be in.
• The OOP paradigm gives you the option to group
parts of the state into objects, thus making it easier
to manage.
• But it doesn’t significantly reduce the number of
possible states.
Why mutable state arises??
This example needs to modify a couple of independent variables
while counting the number of newlines in a single file. Some changes
depend on each other, and others don’t.
If you allow the user to modify the text while you’re
saving it, incomplete or invalid data could be saved, thus
creating a corrupted file.
If you either create a full copy or use a structure that can remember
multiple versions of data at the same time, you can decouple the
processes of saving the file and changing the text in the text editor.
Recap:
How to reduce mutable state.
for (const auto& file: files) {
int line_count = 0;
std::ifstream in(file);
while (in.get(c)) {
if (c == '\n') {
Avoiding line_count++;
mutable }
state (1) }
results.push_back(line_count);
}
int count_lines(const std::string& filename)
{
std::ifstream in(filename);
return std::count(
std::istreambuf_iterator<char>(in),
Avoiding std::istreambuf_iterator<char>(),
mutable '\n');
state (2) }
std::vector<int>
count_lines_in_files(const std::vector<std::string>& files)
Avoiding {
mutable return files | transform(count_lines);
state (3) }
Thinking Functionally
• It would be inefficient and counterproductive to
write code in the imperative style first and then
change it bit by bit until it became functional.
• Instead of thinking of the algorithm steps, you
should consider what the input is, what the output
is, and which transformations you should perform to
map one to the other.
When thinking functionally, you consider the
transformations you need to apply to the given
input to get the desired output as the result.
You can perform the same transformation on each
element in a collection. This allows you to look at the
simpler problem of transforming a single item instead of
a collection of items
You can decompose a bigger problem of counting the
number of lines in a file whose name you have into two
smaller problems: opening a file, given its name; and
counting the number of lines in a given file.
By using transform, you can create functions that can
process collections of items from functions that can
process only one item at a time.
Benefits of functional
programming
• Code brevity and readability
• Concurrency and synchronization
• Continuous optimization
• Evolution of C++ as a functional programming
language
• Many more
Benefits of functional
programming
Summary (1)
• The main philosophy of functional programming is that
you shouldn’t concern yourself with the way something
should work, but rather with what it should do.
• Both approaches—functional programming and object-
oriented programming—have a lot to offer. You should
know when to use one, when to use the other, and when
to combine them.
• C++ is a multiparadigm programming language you can
use to write programs in various styles—procedural,
object-oriented, and functional—and combine those
styles with generic programming.
Summary (2)
• Functional programming goes hand-in-hand with
generic programming, especially in C++. They both
inspire programmers not to think at the hardware
level, but to move higher.
• Function lifting lets you create functions that
operate on collections of values from functions that
operate only on single values. With function
composition, you can pass a value through a chain
of transformations, and each transformation passes
its result to the next.
Summary (3)
• Avoiding mutable state improves the correctness of
code and removes the need for mutexes in
multithreaded code.
• Thinking functionally means thinking about the
input data and the transformations you need to
perform to get the desired output.
Thanks!
• Reach out to me (kunalk@goa.bits-pilani.ac.in)
and my TA team

You might also like