Functional Programming - Why and How

An introduction to Functional Programming, given during an FPNCR meetup at S&P offices in January 2018.

Transcript

1. Why and How Functional Programming By Anupam Jain
2. About Me •Organiser of FPNCR meetup •Programming professionally for over 12 years •Mostly Javascript and Haskell •Built large production systems with Haskell •Some FOSS libraries - •Web UIs - (https://github.com/ajnsit/concur) •Servers - (https://github.com/ajnsit/wai-routes)
3. What is Functional Programming •Functions Exist! •Functions are not Procedures. •Input and Output •First class
4. What is Functional Programming •Functional •Declarative • Pure • Immutable • Statically Typed • Algebraic Data Types • Lazy • Concurrent • ……
5. What is Functional Programming •Functional •Declarative • Pure • Immutable • Statically Typed • Algebraic Data Types • Lazy • Concurrent • …… Functions are the mode of abstraction
6. What is Functional Programming •Functional •Declarative •Pure • Immutable • Statically Typed • Algebraic Data Types • Lazy • Concurrent • …… No sequence of steps
7. What is Functional Programming •Functional •Declarative •Pure • Immutable • Statically Typed • Algebraic Data Types • Lazy • Concurrent • …… No side effects
8. What is Functional Programming •Functional •Declarative • Pure •Immutable • Statically Typed • Algebraic Data Types • Lazy • Concurrent • …… No assignments
9. What is Functional Programming •Functional •Declarative • Pure • Immutable •Statically Typed • Algebraic Data Types • Lazy • Concurrent • …… Potentially invalid programs are not compiled
10. What is Functional Programming •Functional •Declarative • Pure • Immutable • Statically Typed •Algebraic Data Types • Lazy • Concurrent • …… User defined data types are sums of products
11. What is Functional Programming •Functional •Declarative •Pure • Immutable • Statically Typed • Algebraic Data Types •Lazy • Concurrent • …… Order of evaluation is unknown
12. The Sweet Spot Productivity Quality Resilience FP
13. Productivity • Steep learning curve. But.. • Abstraction • Types • Isolation • APIs - Immutability, Laziness, Composition
14. Productivity • Very few FP practitioners. But… • Supply vs Demand • High Quality Talent Pool • High Employee Satisfaction • Remote • Training
15. Quality • Excellent Abstractions Pre-cooked for you • Excellent libraries for many things. Such as async programming.
16. Quality • Static types hinder quick scripts. But.. • Easy Refactoring, checked at compile time • Slight Hyperbole - “If it compiles it works” • Reality - “Refactor without fear” • Let the computer help you help yourself • All this implies - Low Tech Debt
17. Resilience • Forced to Manage Complexity • No Hidden State or Effects • Large class of bugs caught at Compile Time. • Great Refactoring Story.
18. Functions • Functions exist! • Local and anonymous function definitions. • Higher order functions • First class functions
19. Other things • Do notation. General purpose. • Alternative is a natural fit for many domains like parsing, gui construction. You can define many algorithmic functions on top like optional, many, foldMap etc. Almost forms a DSL. • Advanced techniques • Free Monads • Monad Transformers • Lens, Conduit, Pipes
20. Via Haskell • Haskell Syntax • Haskell Semantics
21. FPNCR •The Delhi/NCR Haskell and other Functional Programming Languages Meetup Group •https://www.meetup.com/DelhiNCR-Haskell-And- Functional-Programming-Languages-Group •Github- https://github.com/fpncr •Mailing List - https://groups.google.com/forum/#!forum/ fpncr •Slack - https://functionalprogramming.slack.com - #FPNCR